Loading...
HomeMy WebLinkAboutKake Petersburg Intertie Project Underground Transmission Line Alternative Phase I 1983KAKE-PETERSBURG INTERTIE PROJECT UNDERGROUND TRANSMISSION LINE ALTERNATIVE PHASE I - PRELIMINARY TECHNICAL ANALYSIS REPORT PREPARED FOR THE ALASKA POWER AUTHORITY BY EBASCO SERVICES INCORPORATED JULY 29, 1983 TABLE OF CONTENTS Page 1.0 INTRODUCTION. .......... a er - 1-1 1.1 BACKGROUND ... 2... 2.2.2... eee eee eee 1-1 1.2 PURPOSE OF PRIMARY TECHNICAL STUDIES ....... 1-2 2.0 TECHNICAL FINDINGS... 1... 2. eee ee ee ee 2-1 2.1 ISSUES 2... ew ee ee ee eee ee ee 2-1 2.2 CALCULATIONS ... 1... 1... ee ee ee ee ee 2-2 2.2.1 Voltage Profiles and Loss Estimates .... 2-5 2.2.2 Energizing of the Cable .......... 2-23 2.2.3 Outage Analysis . 2... 2.2.2 ee ee eee 2-24 3.0 SUMMARY AND RECOMMENDATIONS ........2..2006. 3-1 4040B Vi Number 2-1 2-2 LIST OF TABLES Title CIRCUIT PARAMETER COMPARISONS USING MEDIUM LINE (M) AND LONG LINE(L) FORMULAS SUMMARY OF VOLTAGE VARIATIONS AND LOSSES Vii Page 2-4 2-7 Number 2-1 2-2 2-3 2-5 2-6 2-7 2-8 2-9 2-10 2-10 LIST OF FIGURES Title CIRCUIT PARAMETERS PER MILE LENGTH OF CABLE. 1/0 AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. NO COMPENSATION. NO LOAD AT KAKE. 1/0 AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. NO COMPENSATION. FULL LOAD AT KAKE. 1/0 AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. 2 MVAR COMPENSATION AT BUS #4. NO LOAD AT KAKE. 1/0 AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. 2 MVAR COMPENSATION AT BUS #4. FULL LOAD AT KAKE. 500 kCM AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. NO COMPENSATION. NO LOAD AT KAKE. 500 kCM AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. NO COMPENSATION. FULL LOAD AT KAKE. 500 kCM AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. 3 MVAR COMPENSATION AT BUS #4. NO LOAD AT KAKE. 500 kCM AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. 3 MVAR COMPENSATION AT BUS #4. FULL LOAD AT KAKE. 500 kCM AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. 2.5 MVAR COMPENSATION AT BUS #2 and BUS #4. NO LOAD AT KAKE. 500 kCM AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. 2.5 MVAR COMPENSATION AT BUS #2 and BUS #4. FULL LOAD AT KAKE. iv 2-9 2-10 2-11 2-12 2-13 2-14 2-15 2-16 2-17 Number 2-12 2-13 2-14 2-15 LIST OF FIGURES (Continued) Title 500 kCM AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. 0.75 MVAR AT ALL BUS POINTS. NO LOAD AT KAKE. 500 kCM AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOL TAGE. 0.75 MVAR AT BUSES #2 TO #6. FULL LOAD AT KAKE. 750 kCM AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. NO COMPENSATION. NO LOAD AT KAKE. 750 kCM AL CONDUCTOR, 260 MIL XLP INSULATED, 25 kV NOMINAL VOLTAGE. NO COMPENSATION. FULL LOAD AT KAKE. Page 2-18 2-19 2-20 2-21 1.0 INTRODUCTION _ 1.1 BACKGROUND Ebasco Services Incorporated (Ebasco) submitted a Draft Feasibility Report for the Kake-Petersburg Intertie Project to the Alaska Power Authority (Power Authority) in November of 1982. Public meetings on the report were held in December of 1982 and agency and public comments were received through February of 1983. Some of the comments received suggested that it would be advisable to consider alternatives involving longer segments of underground cable, as well as a different route along the north corridor as identified in the projects Routing and Environmental Report. After the close of the comment period, a meeting was held in March, 1983 involving the Power Authority staff, Ebasco personnel, and a representative of Dryden and LaRue Engineering, a firm with experience in a project involving a long underground cable installation. During the March meeting, the merit of proposing an underground option for the Kake-Petersburg Project was discussed. A Summary of that meeting is provided in Appendix A. As a result of the March meeting, it was decided that additional Studies to explore the viability of constructing an underground transmission line between Kake and Petersburg were warranted. It was also determined that the most efficient and logical way to study underground and new routing options would be to conduct additional Studies in a step-wise manner. The initial step would involve an assessment of the technical viability of an underground line without considering route-specific factors. That approach was adopted because only if voltage drop, excessive line charging, and other such Considerations could be kept to acceptable levels would it be worthwhile to consider the underground option. Further, it was recognized that the costs associated with installing an underground cable could be attractive, depending on site specific conditions. Thus, preliminary supplemental studies were authorized to address the factors which will determine if more detailed reconnaissance level, route-specific studies are warranted. This report addresses such considerations. 1-1 1.2 PURPOSE OF PRELIMINARY TECHNICAL STUDIES The preliminary technical studies described in this Phase I report are designed to look at the electrical design parameters influencing the proposed underground option. Factors requiring analysis include analysis of voltage fluctuation, whether or not compensation would be required, and whether the proposed underground cable would be compatible with the existing and proposed systems in Kake and Petersburg. The expected failure rate also merits consideration. Information on failure rates is needed to be able to estimate average outage times for repairing the line. The quantification of failure rates requires analysis of industry experience with underground Cables. Another important consideration, cost, is not addressed in this report. Cost considerations will be dealt with later, when more route-specific data become available. 4040B 2.0 TECHNICAL FINDINGS Ebasco's investigations regarding the technical feasibility of constructing an underground line from Kake to Petersburg did not locate any other projects which included an underground cable transmission line approximately 50 miles long at voltage levels between 13.2 kV and 35 kV. Such lines were considered in the past to be unfeasible because it was assumed that the charging currents and the voltage rise at the end of the cable would be excessive. Furthermore, the cost of the cable and its installation was noncompetitive when compared to overhead transmission, In the past decade, solid dielectric cables for the previously mentioned voltage levels became commonplace and their cost came down considerably. Direct burying of cables became an everyday practice resulting in total installation costs comparable to that of overhead lines. Such installations, however, have only been developed by utilities serving relatively densely populated areas. In the present analysis, the experience gained on such cable installations is being applied to the Kake-Petersburg project. 2.1 ISSUES In this Phase I study, the investigation was conducted without regard to specific routing considerations. The length of the line between Kake and Petersburg was assumed to be 50 miles and the load to be transmitted was 1.6 MW at a power factor of .9, which corresponds to a 1.8 MVA load and a reactive load of 0.82 MVAR. For this engineering analysis four major items were investigated: ° The electrical engineering feasibility of such a line, i.e., whether or not it can be designed at all. If the line can be designed, what degree of compensation would be required? 4043B 2-1 0 The anticipated effects of energizing the underground cable transmission line on the operation of the Tyee Lake System. ° Estimation of the expected failure rate of such a cable, based on industry data and the anticipated mean outage time due to failures. 2.2 CALCULATIONS Twenty five (25) kV rated voltage cables were selected in the study with 260 mil XLP (crosslinked polyethylene) insulation. The smallest conductor size considered was 1/0 AWG aluminum, because it was felt that this is the minimum size which is capable of carrying the 1.8 MVA load at 24.9 kV. In addition, 500 kCM and 750 kCM conductor sizes were also evaluated. The 50-mile length of underground cable was divided into five, 10-mile long segments. Equivalent pi representation was used. The circuit parameters per unit length are shown in Figure 2-1. Calculations were made to verify whether the 10-mile segment lengths would give reasonably good approximation. In Table 2-1, comparative data are given for 10-mile and 50-mile line lengths. For each of the conductor sizes, the equivalent pi values are calculated using the medium long line method and the long transmission line formulas. In the medium long line method, the per-mile parameters shown in Figure 2-1 were multiplied by the appropriate line lengths. The long transmission line values were calculated using the hyperbolic sine and tangent functions. The results indicate that the selection of a 10-mile segment does not introduce any appreciable errors into the calculations. 1/ 0. I. Elgerd, Electric Energy Systems Theory: An Introduction, meceay Hill, New York, 1971, pp. 184 and 190: Equations (6-72) and -94). 2-2 R(ohm/mile) X(ohm/mile) C(microfarad/mile) 1/0 AWG 1.1 0.49 0.24 500 kCM 0.24 0.4 0.45 750 kCM 0.16 0.38 0.465 Aluminum conductor, 260 mil XLP insulated. FIG 2-1 Circuit Parameters Per Mile Length of Cable. TABLE 2-1 CIRCUIT PARAMETER COMPARISONS USING MEDIUM LINE (M) AND LONG LINE (L) FORMULAS!/ 2/ R nN 10.9 2.4 2.4 1.6 1.6 x, 4.9 4.9 4.0 4.0 3.8 3.8 XeV/ 1,100 ‘1,098 590 587 570 568 50 Miles R 55 52.9 12 11.3 8 7.56 x, 24.5 26.3 20 19.6 19 18.7 ov 220 217 118 1 114 108 Y/ Both methods use pi equivalent circuits. (M) multiplies/divides the per mile parameter values by the number of miles, but (L) uses formulas with hyperbolic sine and tangent functions to calculate the parameters of the equivalent circuit. 2/ The reactance of Xc corresponds to the capacitance of each end of the pi equivalent, i.e., to the value C/2 in Figure 2-1. eSSSSSSSSSSSSSSSSSSSMFFseF 2-4 2.2.1 Voltage Profiles and Loss Estimates The voltage profiles and loss calculations were carried out using a TI-59 programmable calculator with a PC-100C printer. Results were verified using the Westinghouse "Westcat" load flow program. The runs on the Westinghouse computer proved the correctness of the results obtained using the TI-59. In fact, it turned out that the TI-59 will provide more accurate results because it works with 13 decimal-digit accuracy and prints out 10-decimal digits. The input data for the load flow program of the TI-59 calculator were the network parameters, and the load and voltage at the receiving end. These were chosen because of the limitations imposed by the memory size of the calculator. This means that the data are input at the receiving end from which the data for the sending end are calculated. Consequently, the computations were carried out by starting at Kake and moving, so to speak, backward towards Petersburg. This is the reason why in all the calculations, the voltage at Kake was maintained at 24.9 kV. As the voltage drops are not significant in any of the cases, the results are indicative and no further iteration was done to calculate the voltage and loss conditions when the voltage is 24.9 kV at Petersburg. It was assumed in the loss calculations that the losses of the reactors are 1% of their MVA rating. This is probably on the high side. At this phase of studies, however, we considered this assumption to be adequate. Apparently, there is at least one firm willing to deliver reactors in sizes of 3 MVAR or smaller. Should it turn out that air core reactors are available in these small sizes, the reactor losses could be considerably less. However, this refinement is not needed until more detailed studies are performed (the Economic Analysis of Task III). The results of the computations are summarized in Table 2-2 and presented in detail in Figures 2-2 through 2-15. 4043B Results for the 1/0 Aluminum Conductor: Figure 2-2 shows the no-load case with 1/0 aluminum conductor. It can be seen that the voltage Profile is acceptable, since the voltage rise is only around 5% at Kake. The no-load losses of the line are 210 kW. Figure 2-3 shows the Same line configuration with full load at Kake. The voltage drop at Kake is approximately 12%, which though quite high may still be Compensated with the LTC at Kake. More serious are the losses in this case which are 980 kW or 61.2% of the 1.6 MW load at Kake. In order to try to reduce the losses, Compensation was attempted by placing a 2 MVAR reactor at bus No. 4, or approximately 20 miles out of Kake. The no-load case is shown in Figure 2-4 and the full load case in Figure 2-5. Comparing these figures with the previous ones, it can be seen that the no-load losses have been reduced to about one-quarter of the uncompensated value; however, the losses at full load have been increased compared to the uncompensated case. Because of the large losses, no further studies were made with 1/0 aluminum conductor. Results for the 500 KCM Aluminum Conductor: The next set of figures (Figures 2-6 through 2-13) show the results of studies made on 500 kCM aluminum conductor. The uncompensated cases for the 500 KCM conductor are shown on Figures 2-6 and 2-7, the former being the no load case and the latter the full-load case. The voltage rise at no load is only 9% and at full load it is even less at 3%. The losses are 150 kW at no Toad and 180 kW at full load. The latter represents 11.2% of the 1.6 MW load at Kake. Figures 2-8 and 2-9 show the load flow with one 3 MVAR reactor installed at bus No. 4, or approximately 20 miles out of Kake. The voltage variation between full load and no load remains in the + 3% range at Kake. The no-load losses have increased to 260 kW as compared with 150 kW for the uncompensated value. Also, the losses for full load increased by 40 kW when compared to the noncompensated case. Here the difference is due to the estimated 1% of the 3 MVAR reactor losses or 30 kW. As mentioned earlier, this may be considerably less if air core reactors are used reducing the calculated losses to within 10 kW of the value shown in the uncompensated case of Figure 2-7. 4043B : -6 TABLE 2-2 SUMMARY OF VOLTAGE VARIATIONS AND LOSSES Maximum Maximum Voltage Voltage Span of Total Losses!/ Rise at Drop at Voltage At No At Full Kake Kake Variations Load Load Conductor Compensation % % % kW KW 1/0 AWG None 5 12 17 210 980 1/0 AWG 1 x 2 MVAR 1 16 17 45 1,070 500 kCM None 9 -- 6 150 180 500 kCM 1 x 3 MVAR 3 3 6 260 220 500 kCM 2 x 2.5 MVAR 2 4 6 65 220 500 kCM 6 x 0.75 MVAR 1 4 5 55 295 750 kCM None 9 -- 5 110 100 / — Included are 1% of MVAR rating of the compensating reactors; lower reactor losses could reduce losses significantly in certain cases. si Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Petersburg 24.9 kV ®@ Line losses: 210 kW 23.69(1.00) 1% of compensation MVAR: 0 kW 4 + 2.52 Est. total losses: 210 kw { 2.11 24.08(1.02) : 1.63 24.42(1.03) ! 1.11 24.68(1.04) ! 0.56 24.85(1.05) Kake 24.9 kV © Os FIG. 2-2 1/0 Al conductor, 260 mil XLP insulated, 25kV nominal voltage. No compensation. No load at Kake. 2-8 Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Petersburg 24.9 kV G) Line losses: — 980 kW 28 .35(1.00) 1% of compensation MVAR: 0 kW 2-8 dese Est. total losses: 980 kW @ fis 27.73(0.98) 4 0.94 27.10(0.96) { 0.31 26.44(0.93) { 0.26 25.71(0.91) Kake 24.9 kV © 1.6f 24.90(0.88) ; 0.82 FIG. 2-3 1/0 AL conductor, 260 mil XLP insulated, 25kV nominal voltage. No compensation. Full load at Kake. 2-9 Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Petersburg 24.9 kV Gj) Line losses: 25 kW 24.74(1.00) 1% of compensation MVAR: 20 kW oe fo.78 Est. total losses: 45 kW @® ¢ 0.23 24.83(1.00) { 0.33 24.81(1.00) ® } 0.89 24.68(1.00) 2 MVAR 24.85(1.00) 24.90(1.01) FIG. 2-4 1/0 Al conductor, 260 mil XLP insulated, 25 nominal voltage. 2 MVAR compensation at bus #4. No load at Kake. 2-10 Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Line losses: 1050 kW 29.60(1.00) 1% of compensation MVAR: 20 kW 2.65 | } 0.10 Est. total losses: 1070 kW + 0.67 28.53(0.96) t 1.19 27.49(0.93) { 1.69 26.44(0.89) { 0.26 25.71(0.87) Petersburg 24.9 kV @ 2 MVAR Kake 24.9 kV © am 24.90(0.84) t 0.82 FIG. 2-5 1/0 Al conductor, 260 mil XLP insulated, 25kV nominal voltage. 2 MVAR compensation at bus #4. Full load at Kake. 2-11 Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Petersburg 24.9 kV @ Line losses: 150 kW 22.90(1.00) 1% of compensation MVAR: 0 kW o5) $444 Est. total losses: 150 kW @ : 3.82 23.59(1.03) { 3.00 24.15(1.05) 4 2.06 24.56(1.07) { 1.05 24.82(1.08) Kake 24.9 kV © a 7 FIG. 2-6 500 kCM Al conductor, 260 mil XLP insulated, 25kV nominal voltage. No compensation. No load at Kake. 2-12 Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Petersburg 24.9 kV @ Line losses: 180 kW 24.29(1.00) 1% of compensation MVAR: 0 kw Est. total losses: 180 x 1.704 tas : @ 4 3.29 24.72(1.02) 4 2.31 25.01(1.03) 4 1.27 25.14(1.03) 4 0.22 25.10(1.03) Kake 24.9 kV © 1.64 {0-82 24.90(1.03) FIG. 2-7 500 kCM Al conductor, 260 mil XLP insulated, 25kV nominal voltage. No compensation. Full load at Kake. 2-13 Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Petersburg 24.9 kV (}) Line losses: 230 kW 24.27(1.00) 1% of compensation MVAR:_ 30_kW ae fe.12 Est. total losses: 260 kW @® 4 1.35 24.53(1.01) -63(1.01) -56(1.01) .82(1.02) 24.90(1.03) FIG. 2-8 500 kCM Al conductor, 260 mil XLP insulated, 25kV nominal voltage. 3 MVAR compensation at bus #4. No load at Kake. 2-14 Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Petersburg 24.9 kV (}) . Line losses: 190 kW 25.70(1.00) 1% of compensation MVAR: 30_kW 1.79f fits Est. total losses: 220 kW @ { 0.37 25.67(1.00) { 0.69 25.49(0.99) { 1.73 25.14(0.98) { 0.22 25.10(0.98) 3 MVAR Kake 24.9 kV © oy sotto { 0.82 FIG. 2-9 500 kCM Al conductor, 260 mil XLP insulated, 25kV nominal voltage. 3 MVAR compensation at bus #4 Full load at Kake. 2-15 Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Line losses: 15 kW 24.43(1.00) 1% of compensation MVAR: _50_ kw 0.02 | + 0.1 Est. total losses: 65 kW @ + 0.90 24.37(1.00) Petersburg 24.9 kV ®@ 2.5 MVAR @ { 0.60 24.55(1.00) { 0.44 24.56(1.01) { 1.05 24 .82(1.02) 2.5 MVAR Kake 24.9 kV © seal FIG. 2-10 500 kCM Al conductor, 260 mil XLP insulated, 25kV nominal voltage. 2.5 MVAR compensation at bus #2 and bus #4. No load at Kake. 2-16 Tyee Lake System Petersburg 24.9 kV @ 25.85(1.00) @ ; 1.62 25.52(0.99) { 0.18 25.41(0.98) 4 1.23 25.14(0.97) 4 0.22 25.10(0.97) 2.5 MVAR 2.5 MVAR Kake 24.9 kV © a Mu ; 0.82 FIG. 2-11 Voltages in kV and (p.u.) Loads in MW/MVAR Line losses: 1% of compensation MVAR: Est. total losses: 500 kCM AL conductor, 260 mil XLP insulated, 25kV nominal voltage. 2.5 MVAR compensation at bus #2 and bus #4. Full load at Kake. 2-17 170 kW 50_kW 220 kW Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Petersburg 24.9 VQ Line losses: 10 kW 24.60(1.00) 1% of compensation MVAR: 45 kW 0.75 MVAR 0.0 fia Est. total losses: 55 kW @ f 0.45 <YY 24.75(1.01) 0.75 MVAR { 0.16 24.86(1.01) 0.75 MVAR ® { 0.14 24.92(1.01) 0.75 MVAR © | 0.45 24.94(1.01) 0.75 MVAR kake 24.9 kV 6 24.90(1.01) 0.75 MVAR gg | 40.75 FIG. 2-12 500 kCM Al conductor, 260 mil XLP insulated, 25kV nominal voltage. 0.75 MVAR at all bus points. No load at Kake. 2-18 Tyee Lake System Petersburg 24.9 kV 0.75 MVAR i @ $0.7 + 0.35 25.92(0.99) 0.75 MVAR } 0.67 25.73(0.99) 0.75 MVAR 0.98 25.50(0.98) t 1.28 25.22(0.97) 0.75 MVAR © 0.75 MVAR Kake 24.9 © 75 0 MVAR 1.6 f 1.87 FIG. 2-13 26.07(1.00) 24.90(0.96) Voltages in kV and (p.u.) Loads in MW/MVAR Line losses: 250 kW 1% of compensation MVAR: 45 kW Est. total losses: 295 kW 500 kCM Al conductor, 260 mil XLP insulated, 25kV nominal voltage. 0.75 MVAR compensation at buses #2 to #6. Full load at Kake. 2-19 Tyee Lake System Voltages in kV and (p.u.) Loads in MW/MVAR Line losses: 110 kW 22.93(1.00) 1% of compensation MVAR: 0 kW oud + 4.60 Est. total losses: 110 kW Petersburg 24.9 kV ®@ 23.61(1.03) { 3.11 24.16(1.05) { 2.13 24.57(1.07) 4 1.08 24.82(1.08) Kake 24.9 kV © TN FIG. 2-14 750 kCM Al conductor, 260 mil XLP insulated, 25kV nominal voltage. No compensation. No load at Kake. 2-20 Tyee Lake System Petersburg 24.9 kV @ Kake 24.9 kV © 1.6 { FIG. 2-15 26 39 24 39 24 {0.82 24.04(1.00) .51(1.02) .85(1.03) .03(1.04) .05(1.04) 24.90(1.04) Voltages in kV and (p.u Loads in MW/MVAR Line losses: 1% of compensation MVAR Est. total losses: 750 kCM Al conductor, 260 mil XLP insulated, 25kV nominal voltage. No compensation. Full load at Kake. 2-21 -) 100 kW kW *__0_ kW 100 kW Figures 2-10 and 2-11 show the conditions with two 2.5 MVAR reactors installed at bus No. 2 and 4, which are located approximately 10 miles out of Petersburg and 20 miles out of Kake, respectively. The voltage variation is again close to the + 3% range and the voltage profile is very good. At no load line losses are small, only 15 kW; however, the estimated 1% of the two, 2.5 MVAR reactors adds 50 kW to the losses resulting in a total loss of 65 kW. At full load, the losses are reduced to 220 kW or 13.8% of the Kake lodd. Should supplier(s) indicate that air core reactors of this size can be delivered and have a considerably less loss, the condition would improve considerably. In this case, for example, the losses at full load may be as low as 10.6%. An evenly distributed compensation was also attempted. In this case, each of the six buses were furnished with 0.75 MVAR reactors as shown in Figures 2-12 and 2-13. The voltage profile was very good, however, the losses at full load are 295 kW and thus are not as good as when only two, 2.5 MVAR reactors were employed. Results for the 750 kCM Aluminum Conductor: Finally, 750 kCM aluminum conductor was also investigated. Figures 2-14 and 2-15 show the results. The voltage rise at no load and full load is approximately 9% and 4% at Kake, respectively. The no load losses are 110 kW, that is 40 kW less than for the uncompensated 500 kCM cable line but still not as favorable as the two reactor compensated versions of the 500 kCM line. However, losses at full load are the lowest in this case, even if one assumes that the compensating reactors have no losses in the other cases. Conclusion of Voltage Profile and Losses: The analysis of voltage profiles for the 1/0 AWG, 500 KCM, and 750 KCM reveals that there are a range of acceptable solutions for an underground cable transmission line between Petersburg and Kake. The analysis of the alternatives Show that all of the conductors considered could work, but that identifying the optimal solution requires a trade off analysis of the Cost of the conductor, line losses, and the required compensation. For 2-22 example, selection of the 1/0 AWG conductor, which is the least expensive to install, leads to the most line losses. Further, if the 1/0 AWG conductor is selected and the value of the line losses is cosidered negligible the 1/0 AWG conductor is the best choice because there would be no dollar value attributable to the losses. Consequently, the conductor selection problem becomes mainly a function of how it is decided to treat line losses. Once a dollar value is assigned to the losses, selection of the optimal conductor and level of compensation can be made. 2.2.2 Energizing of the Cable The affects of energizing the line on the 24.9 kV bus in Petersburg was estimated based on information developed in the Tyee Lake System studies.2/ The basis of the calculation are the results shown in Figures III-B6 and III-B7 of that report. In both figures, the bus at Petersburg is held at 24.9 kV while the reactive power increases from 1.8 to 4.1 MVAR, an increment of 2.3 MVAR. During this change in reactive power, the low side per unit taps moved from 1.05 to 1.13 at Petersburg. This means that 2.3 MVAR will cause approximately 8% change in voltage on the 24.9 kV bus in Petersburg. Whether a 500 kCM or a 750 kCM cable is needed, the maximum line charging at Petersburg is around 5 MVAR at 24.9 kV. Using the above figures, the steady state voltage rise at Petersburg will be around 17%, for an uncompensated cable. However, should the cable be fully compensated, then practically no voltage rise will occur at Petersburg during energizing. 2/ Tyee Lake-Wrangell-Petersburg Power Systems Study, Final Report, Ebasco, October 8, 1982. 2-23 The preceding calculations are based on steady state values. No attempt was made to assess above 60 Hz oscillations or any other effects that may be caused by energizing the cable system. Such detailed studies will be conducted under Optional Task III-C, if necessary. 2.2.3 Outage Analysis In this section a summary of the analysis of the expected outage rates for the Kake-Petersburg cable alternative is given, Unfortunately, there is very little data available on actual cable failures. In spite of this lack of data and the fact that utility companies, trade associations and manufacturers are unwilling to divulge details, Ebasco was able to assemble information from which reasonable assessments could be made regarding the expected failure rates and associated repair costs for the Kake-Petersburg cable system. Cable experts agree that most cable failures are caused by "dig-ins" or splice failures. Actual cable failures caused by treeing are few. From statistical data available within the Pacific Northwest, it can be concluded that the expected mean time between failures should be approximately five years for a 50-mile long three XLP-insulated cable line. The mean time between failures should be even longer for Southeast Alaska than it is for the Pacific Northwest. Evidence obtained on failures to date seem to indicate that more cables fail in the summer than in the winter. As the summer temperatures are much higher in almost al] of the Pacific Northwest than in Southeast Alaska, fewer failures can be expected in the Kake-Petersburg area than were experienced in the Pacific Northwest. The milder winter temperatures of Southeast Alaska should help in the right direction also. 2-24 Furthermore, recently installed XLP-insulated cables have extremely good reliability records though there is not sufficiently long operating experience with such underground cable installations to reach a definite conclusion. Many cable engineers feel that these cables will have trouble-free operation for most of their 50-year expected life. However, it must be stressed that the trouble-free operation of the cable hinges on two facts: strict quality control of the cable before installation and strict adherance to high-quality splicing techniques. If failures, particularly at splicings, do not occur within a few days of operation of the cable, the chances are very good that failures will not happen later, unless “dig-ins" occur. In the case of the Kake-Petersburg line "“dig-ins" appear unlikely. Therefore, one can expect that the mean time between failures will be more than the five years given above, because that finding is based on eight years of operating experience involving old XLP-insulated cables. The length of time the underground cable line would likely be out of service in the event of a failure depends on when and where the failure occurs. If a failure occurs in the winter, it could be up to several months until conditions would permit repair of the cable. In the summer it is likely that repair could be accomplished more readily. The remoteness of the area where the cable is located also influences the amount of time until repair could be made and the associated cost. In general, if helicopters must be used to transport workers and equipment, costs will be high. Recognizing that conditions vary considerably over any route from Petersburg to Kake, Ebasco's studies Suggest that repairing the typical cable failure in an unroaded area would cost between $11,000 and $25,000, depending on weather and other factors. If four cables are installed, the outage time should be minimal because switching in the spare cable could be accomplished in a very short time. Considering that even under adverse conditions repair could probably be accomplished in less than one month, the likelihood of a 2-25 second failure occurring during the same time period is very low. Therefore, the outage time due to a cable failure can be minimized to about one-half hour, the time it would take to get linemen out to the substations and do the appropriate switching. Should the process be automated, which can be done easily, the outage time can be reduced to the order of one minute. Locating a fault should cause no problems. Recently, one of the manufacturers introduced on the market a cable fault locator (radar) with a range up to 100 miles. This means that the approximate location of the fault can be established from any end of the cable system. To facilitate the finding of the exact location of a fault, risers are envisioned at approximately every mile. The lowest range of the equipment mentioned is 100 feet. With a four cable arrangement, even two faults can be tolerated, provided that the two faults do not occur within the same two risers. As mentioned earlier, the common occurrence of two faults is very unlikely, therefore the occurrence of two faults between the same two risers is even more remote. 2-26 3.0 SUMMARY AND RECOMMENDATIONS NAT TONS The preliminary analysis of the underground option indicates that constructing an underground cable is technically viable. The selection of voltage and size of the cable and type of compensation depends on a tradeoff analysis of cable and reactor costs and line and reactor losses. Installation costs based on route specific considerations will, of course, also influence the facilities to be proposed. Such Studies will be conducted in the reconnaissance (Phase II) studies if authorized by the Power Authority. Along with analyzing route specific considerations, an economic analysis of line losses and outage related consideratrions is needed in the Phase II studies. The likelihood and cost of outages occurring as a result of cable problems in remote areas during winter months, when Corrective measures would be very difficult and expensive, could be analyzed if so desired, although the findings indicate that such costs would be acceptable, In order to get a basis for economic evaluation, it is recommended that the APA establish a policy regarding the value of line losses, while transporting power obtained from the Tyee Lake Project, prior to proceeding with the economic evaluation of the reconnaissance-level Studies of the underground option. The costs associated with losses may considerably influence the size of cable ultimately selected and the selection of other basic project parameters. Until now, the value assigned to line losses in the feasibility analyses was taken zero, as there was no additional cost associated with the increment of power to be lost; the availability of Capacity at the Tyee Lake project provided the basis for this assumption. The loss accounting policy should be finalized by APA before proceeding with the economic evaluation of a selected cable transmission system. 4045B 3-1 APPENDIX A KAKE-PETERSBURG INTERTIE PROJECT Meeting Summary - March 3, 1983 Additional Studies Related to Underground and North Corridor Options for the Proposed Kake-Petersburg Project Attendees Miles Yerkes, Alaska Power Authority Patty DeJong, Alaska Power Authority Remy Williams, Alaska Power Authority William Kitto, Ebasco Services Inc. John Szablya, Ebasco Services Inc. Malcolm Menzies, R & M Consultants, Inc. Robert Dryden, Dryden and LaRue Consulting Engineers Meeting Summary On March 3, 1983, a meeting was hela to review the advisability of conducting additional studies to analyze new alternatives for constructing the proposed Kake-Petersburg transmission line. New alternatives discussed included a line in either the north corridor or a line using an underground cable for most of its length. The meeting began with Bill Kitto reviewing the status of project activities. He described that a Feasibility Report had been submitted during November, 1982, and that the report had dismissed the possibility of constructing the line underground because of technical and reliability considerations. The report also failed to evaluate the north route in detail because of environmental considerations and because of the fact that it was well removed from any existing access corridors. Following the release of the Draft Feasibility Report, public meetings were held in Kake and in Petersburg and comments were obtained. Bill Kitto explained that the primary opinion expressed at the Kake meeting was A-1 that residents of that community felt that interconnection to the Tyee System would bring them cheap hydroelectric power. This view by the residents of Kake prevailed, although it was repeatedly explained by Remy Williams and Bill Kitto that the Tyee power which would be transmitted to Kake by the Intertie would not be low cost power. Despite these explanations, residents of Kake continued to state the opinion that interconnection with the Tyee System would enable their community to prosper because of low electricity rates and enable them to install electric heat and other large electricity-consuming devices. Bill Kitto also explained that comments were received from the public regarding the load forecast stating that that forecast was too low and did not properly reflect the growth in Kake. Bill Kitto explained that subsequent to that meeting, David Reaume, author of the load forecast, considered these comments and concluded that the initial forecast was indeed accurate and that the most recent growth in electricity consumption was related primarily to new purchases as a result of the permanent fund checks received by residents of the community of Kake. Bill Kitto also explained that two important comments were received on the Draft Feasibility Report. The first comment came from the U.S. Forest Service who indicated that they were currently planning to let contracts on a road system from the Hamilton Creek area east toward Portage Bay. This would reauce the total length of the line between Kake and Petersburg within the north corridor substantially. The second important comment related to the Alaska Lepartment of Transportation and Public Facilities who indicated that they would be interested in participating in a joint effort to construct a transmission line/road corridor to Kake. Based on these two agency comments and the Power Authority's request to look at the underground option in more detail, Ebasco Services was recommending that the Power Authority prepare a brief report to look into the option of constructing an overhead-underground line within either the north or south corridors linking Kake and Petersburg. Bill Kitto explained that A-2 it was the main purpose of this meeting to initiate these studies and to discuss risks and tradeoffs involved in underground construction in such an area. Following Bill Kitto's remarks, Bob Dryden described the work he had been involved in in Iliamna. He explained that this project involved constructing approximately 35 miles of 24.9 kV underground cable. He explained that they had investigated construction of an overhead line in this area, but had discarded it because of economic reasons and because of potential impacts. A particular concern in the Iliamna area was the potential impact an overhead line would have on aircraft. They also considered an underground system because they believed use of a vibratory plow would enable them to install an underground cable economically in the area. Bob explained that there were three types of vibratory plows which they had considered, and that they were very happy with the installation approach they ultimately selected. He suggested that the contractor, Dodge Electric of Wasilla, Alaska, be contacted for specific information about the approach they used. Bob then described several of the important considerations related to their experience in lliamna. In reviewing some of the drawbacks, Bob explained that the installation approach they had used required three operators. He said two cats, a 0450 and a D6, were needed, as well as a vehicle to properly handle and transport the cable. The cats had larger size motors to handle the hydraulics. The cat with the vibratory plow had one cable drum, the other had two drums. The following specific points were made by Bob Dryden and others: 1. Installation costs ran about $3/ft on the Iliamna project. 2. Geotechnical conditions were different at Iliamna than in Southeast Alaska. He said that in particular, the soils in 3. the Iliamna area were largely composed of alluvial material and were relatively easy to work when compared with some of the hardpan found in Southeast Alaska. Bob cautioned that not as much concern needs to be given to the depth of seasonal frost in the Kake area. While discussing this point, Malcolm Menzies pointed out that in some years, the ground did not freeze in Southeast Alaska, depending on when the first heavy snowfall occurred. The cable was buried to a depth of approximately 42 to 48 inches on the lliamna project, although it probably would have been possible to bury it at a shallower aepth. When areas of bedrock were encountered on the Iliamna project, Bob Dryden indicated that they rerouted the underground cable to avoid such areas. The cable used at Iliamna was 25 kV, single phase, URD type, with cross linked polyethelene insulation, 1/0 aluminum conductor and concentric neutral. It had a diameter of 1.45 inches overall. Use of 15 kV cable instead of 25 kV would reduce costs by 1% to 2%, however, the 25 kV is stronger and can take much more of a beating. In describing the Iliamna project, Bob Uryden stated that the length of the cable installed was approximately 35 miles, and that a portion of it had been energized since October, 1982 without any failure. It was explained that the cable installation equipment installed approximately 2,000 feet of cable per hour when A-4 10. 11. 12. 13. working effectively. In general, because of the vibratory nature of the equipment, it was difficult to work more than 2-1/2 to 3 hours per day. As a general rule, workers worked approximately two hours per day and repaired the equipment approximately ten hours per day. On the average they laid 1/2 mile in the morning and 1/2 mile in the afternoon. The line was constructed between August 1 and September 15, 1982. There were approximately 30 phase breaks on the Iliamna project; the cable snapped when the pull was too much. In spite of the number of breaks, a break never went undetected and there was no marginal damage to the cable. In general, Bob Dryden stated that the vibratory plow cannot deal with overburden material even as thin as 2 inches, because it would clog the equipment and keep it from functioning properly. Bob Dryden also recommended that because of the unique operating conditions of the vibratory plow and other cable installation equipment, it would be advisable for the contractor who installed the line at Iliamna to be sent to Southeast Alaska to judge the problems that could be expected from installing such a line in that area. He said that such an on-the-ground inspection would be well worth the money prior to proceeding with the construction of such a line. Bob reviewed a number of ways that his firm had tried to eliminate problems related to the organic layer of the soil which had made their machine inoperable. He explained that using a ripping technique to cut the vegetation prior to installation of the cable was less than satisfactory. He suggested that removal of the material prior to installation of the cable was the most prudent approach, by driving a rolling hoe ahead of the vibratory (or other) cable plow. 14. 15. 16. 7. Electrical charging problems were also identified as a concern and various solutions for such problems were discussed. Although the Iliamna line was not compensated, Bob said he would compensate similar future lines. The use of lightning arrestors was tried initially by Dryden and LaRue, but were found to be unacceptable, instead Ohio Brass MOV type surge arresters were used. Discussions of installation techniques led to the finding that jit might be worthwhile to consider use of a frostwheel on the project and to install the line during winter months. Bob Dryden felt that such an approach would make sense, given the anticipatea problems of using a vibratory plow in the forested, densely vegetated area of Southeast Alaska. Mike Yerkes provided his comments on the feasibility of an underground line and suggested that Ebasco look into various types of wheel operators which could be contacted to determine the costs and practicality of using such an approach in Southeast Alaska. Mike also suggested that future studies determine the parameters affecting the results of the study by considering the length of the cable, size of the cable, reliability of the cable and what happens under faulting. Bob Dryden reported that two faults were experienced on the Iliamna project. One fault remains unexplained, while the other occurred on a cable section routed across a lake where the cable sat on a sandbar and was under the pressure of ice which finally damaged the cable. They used compression-type splices with slip back sleeves and found the ones GE made to be the best. For fault location it is necessary to have trained people and suitable equipment available. Bob explained that it was very necessary to sectionalize the line frequently over its entire length. He suggested using a pad 18. 19. 20. 21. with a ground-galvanized sheath system to install sectionalizing equipment. He said such an approach was essential as it was very undesirable to remove large quantities of frozen ground in order to locate a fault in the line, should one occur. Mike Yerkes indicated that he was concerned about the reliability of the system and said that any future studies which Ebasco would conduct should examine the reliability criteria for such a system. He suggested that a six-week outage option would be reasonable. Bob Dryden said that given the difficulty in locating and repairing an outage in the winter months, it would seem reasonable that any major fault occurring during the winter would make the line inoperable for the rest of the winter. Mike Yerkes recapped his feelings on the need for additional studies by stating that initially studies should focus on an all underground option. He indicated that this was needed so that a determination could be made, whether it was technically feasible to construct an underground line. To do this, an underground-only option, regardless of any routing considerations, would be studied. This technical analysis should consider the technical feasibility (line charging), reliability and costs prior to investigating any site-specific issues. Site-specific factors would be considered later, if warranted. Bob Dryden also commented that it was important to consider proper operating and maintenance costs as maintenance might be very costly for an underground system. He also pointed out again that it was necessary to consider. the fact that the line could not operate as frequently as could an overhead system.