F-2014 EE 456 Design Project PROJECT REPORT MALUWELMENG, CONNIE SHARP, MEGAN
Table of Contents Introduction... 2 Assignment I... 2 Setup... 2 Simulation... 2 Assignment II... 3 Problems... 3 Solution... 3 Assignment III... 3 Adding Load... 3 Resolving System... 4 System Cost... 4 Conclusion... 5 References... 5 Appendix I System Map... 6 Appendix II Impedance... 7 Appendix III Cost Data... 8 Line Costs... 8 Capacitor/Inductor Bank Costs... 8 Appendix IV One-line Diagrams... 9 Assignment I... 9 Assignment II... 10 Assignment III... 11 Appendix V Assignment I Results... 12 Basic Plan... 12 Line 5-11 Down... 13
Introduction For this project, PSS/E 33 was used as the power flow analysis program. Three assignments were necessary for the project; Assignment I was step-by-step basics of PSS/E 33, Assignment II included the design and modification of a basic transformer system with 161kV and 69kV lines, Assignment III required an additional 40MW load bus to be added as well as an overall load increase of 30% to the existing system. Using PSS/E 33, the per unit bus voltages and real and reactive power were simulated for the given system. Through observation and tracing the power flow, necessary modifications were identified, changed, and finally simulated again. This project was an insightful look at how the concepts shown in EE456 (Power Systems Analysis I) can be applied to realistic transmission systems that require improvements and modifications as the needs for power change. Assignment I Setup The first portion of the project consisted of setting up the Eagle Power System in PSS/E 33. Seventeen buses were represented of which three were generators (including a slack bus). The remaining buses were all load buses. Most of the buses in the system had a base power of 161 kilo-volts. Two buses had a base power of 69 kilo-volts. Since all the generators operated at 161kV, the base power of those two buses was achieved by two step-down transformers in the system. Each transformer had a secondary winding of 1.07 turns, impedance of 0.1333 per unit. Then, realistic transmission lines were added to the model. That is, lines had resistance, reactance, and charging capacities. These parameters were based on the type of conductor used, the length of the line, and the base voltages of the buses being connected. After all of the buses were connected, specifics of each of the buses were added. Load data (real and reactive) were added to all of the load buses. Similarly, maximum and minimum values for both real and reactive power were added to the generator buses. They each could generate a maximum of 430MW of real power and reactive power within the range of -100MVAR to 250MVAR. Simulation With the Eagle Power System all modeled in PSS/E, a simulation was run. The simulation used the Full- Newton Raphson method. Just like hand calculations in which there is no specified start, the system was simulated starting with a flat start. The initial simulation just looked at the results without any changes to the system. The resulting voltages and angles for each bus can be seen in Appendix V. Also found there are the real and reactive power flows, currents, and generator power outputs (real and reactive).
With results of the initial simulation observed, a contingency plan was simulated. In this case, it was assumed that the line between buses five and eleven was down for some unknown reason. As was expected, the voltages and real power flows at these two buses changed the most between the two simulations. Bus five had a 1.78kV difference in voltage magnitudes between the two simulations whereas bus eleven had a 1.95kV. Further comparisons showed that bus five was essentially absorbing real power from bus 11. With this connection gone, bus five started to absorb more from bus eight (double the amount in the initial simulation). Because bus eleven did not have to feed bus five any more, it absorbed less from generators one and two in the second scenario. Assignment II Problems Once the basic case was set up and simulated in PSS/E, we found that not all of our bus voltage values fell within the desired 0.96~1.04 per unit range. Bus 4, bus 5, and bus 7 all had voltages slightly below 0.96 per unit. After examining the system map, we realized that these buses were all in the urban areas and determined that adding a 100MVAR capacitor bank to bus 5 may be a solution. In a couple of the contingency cases (both when Line 10-17 and when Line 13-16 were simulated as offline), the bus voltages for bus 10 and bus 13 were too low. When capacitor banks were added to improve the power factor, they fixed this issue but caused bus voltages for surrounding buses, such as buses 16 or 17, to jump to a higher voltage (somewhere around 1.055 per unit.) Solution After adding the 100MVAR capacitor bank mentioned above into our simulation, it was determined that this fixed all of the issues with low bus voltages in the basic case. It also fixed many similar issues in contingency cases for buses in the urban area of the system. However, there was still the issue of the low bus voltages in buses 10 and 13 when Line 10-17 and Line 13-16 are simulated as offline. Since adding the two 15MVAR capacitor banks to buses 10 and 13 made bus voltages for buses 16 and 17 too high, we decided to add two 10MVAR inductor banks to 16 and 17 as well. This would allow the power factor to improve at buses 10 and 13 while keeping the bus voltages within range for buses 16 and 17. Both the two 15MVAR capacitor banks and the two 10MVAR inductor banks were switch banks - meaning that they are only live for the simulation when Line 10-17 and Line 13-16 are offline. In all other cases, the only bank that is live is the 100MVAR capacitor bank at bus 5. With these inductor and capacitor banks added, we were able to stay in the desired 0.96~1.04 per unit voltage range for all buses in the basic case and within the emergency 0.90~1.05 per unit voltage for all the contingency cases. Assignment III Adding Load Both of the previous portions were baby steps to the actual design project. With the Eagle Power System stable, additional loads were added to the system. Each existing load bus saw a 30% increase in both real and reactive powers. In addition to these changes, a steel mill was added to the grid. Essentially, it was treated as a sixteenth load bus operating at 40MW with a unity power factor.
A new bus meant new lines. Because the mill was situated near generator two and bus fourteen, it seemed a cost efficient idea to just connect the steel mill to these two buses. However, it was decided that running the second line to generator one instead of bus 14 would provide a more stable system overall. The line lengths connecting the mill to generators one and two were 27.2475 and 15.57 miles respectively. This big variance in length led to the decision of connecting to generator two with a Dove conductor (smallest available for a 161kV base voltage). The mill was then connected to generator one with a Drake conductor to account for the longer distance. Resolving System Once all the extra parameters were added to the system, an analysis was done on the system. A basic case was run as well as all the simple contingency plans that had been done for the system before the load change. Based off of these results, it was determined that all the switch capacitor and inductor banks added to the system previously would be live at all times. This change showed that all of the bus voltages for almost all cases (basic and contingencies) fell in the 0.96~1.04 per unit (for the basic) and 0.90~1.05 per unit (for contingencies) ranges desired to keep the system stable. Also, all reactive power flows fell within the limit set with a minimum of -100MVAR and a maximum of 250MVAR. However, there were violations observed when analyzing system results. When Line 10-17 was simulated to be offline, there was a problem with bus 10 having a voltage just below the lower limit. To fix this issue, a 2.5MVAR capacitor bank was switched on once this line went down to increase the power factor at bus ten for this scenario. The basic case in addition to a few of the contingencies also showed voltage issues in the urban areas of the system despite the 100MVAR capacitor bank added previously. Because of this, another 100MVAR capacitor bank was added to bus five. This seemed to stabilize all remaining issues. System Cost To stabilize the initial system, it was decided that the system needed to include three capacitor banks and two inductor banks. Although only given a quote for capacitor banks, it was assumed that costs for inductor banks would be relatively the same. This brought initial costs to $60,000 for each bank for installations. Then, actual banks were priced at $300 per 100kVAR which brought the total cost of stabilizing the system to $750,000. Once the new loads and bus were added, new costs were added. Costs for connecting the steel mill to the grid were incurred along with costs for re-stabilizing the system. Using the Dove and Drake conductors as mentioned in the rural area the steel mill was located led to a cost of $106,000 and $115,000 per mile for the respective conductors. This resulted in a total line cost of about $4.784 million. Lastly, two more capacitor banks were added to stabilize the modified system. With the same cost basis as used for the original system, the additional 100MVAR and 2.5MVAR banks led to a cost of $427,500.
The grand total for all the modifications for the system was approximately $5.961 million as outlined in Appendix II. Conclusion It was decided after the initial system analysis that if the most cost efficient solution would consist of a few capacitor or inductor banks. Line additions were kept to a minimum just because the additional lines connecting the steel mill to the grid proved to be the bulk of the costs incurred for the system modifications. It also made sense to let the voltage base for the steel mill be 161kV based off of its electrical closeness 161kV buses. To give it a 69kV base voltage would lead to installing transformers at one of the nearby buses or routing a line from the existing transformers on the other side of the grid. Both options would have well exceeded the already high costs incurred by simply adding 161kV lines. Without a set systematic way of resolving violations within the system, it was difficult to create any kind of medication. The realization of tracing the power flow to help determine where each bus was obtaining its power from helped in producing better educated guesses as to what should be done stabilize the system. This design project provided a simple but very insightful glance transmission planning. One cannot simply make a change on the grid without seeing how the rest of the system is affected. Because a grid is interlocked, each bus and/or line affects another one way or another. References Bergen, A.R. and V. Vittal, 1999: Power Systems Analysis (2 nd Edition). Colorado State University, n.d.: Introduction to PSS/E. [Available online at http://www.engr.colostate.edu/ece461/labs/lab1_psseintroduction.pdf]
Appendix I System Map 18
Appendix II Impedance Line Conductor Type Resistance 1 Reactance 2 Charging 3 1-9 Drake 3.085 17.47 3.629 1-11 Drake 4.178 26.70 5.550 1-14 Drake 3.629 20.53 4.264 1-18 Drake 2.834 11.66 2.283 2-11 Drake 2.774 15.66 3.251 2-12 Drake 2.618 14.78 3.070 2-14 Drake 3.085 17.47 3.629 2-18 Dove 3.488 10.87 1.685 3-6 Drake 3.551 20.09 4.174 3-12 Drake 3.551 20.09 4.174 3-15 Drake 3.033 17.16 3.569 4-5 Dove 1.529 6.30 1.232 4-9 Drake 2.411 13.69 2.843 5-6 Dove 1.970 8.09 1.584 5-7 Dove 1.089 4.48 0.880 5-8 Dove 1.996 8.17 1.599 5-11 Drake 2.514 14.18 2.949 7-15 Drake 1.866 10.63 2.208 8-12 Drake 1.270 7.13 1.482 10-13 Hawk 3.033 10.15 0.408 10-17 Hawk 3.433 11.49 0.462 13-16 Hawk 4.642 15.54 0.624 1 Resistance is calculated by using R/mile at 50 C in Appendix A8.1 of Bergen and Vittal s Power Systems Analysis. 2 Reactance in the line was calculated by using X/mile in Appendix A8.1 of Bergen and Vittal s Power Systems Analysis. 3 Charging factor was calculated by using (x a+x d)-mile in which x a was obtained from Appendix A8.1 and x d from A8.3 of Bergen and Vittal s Power Systems Analysis. Also taken into consideration was the base voltages of the buses.
Appendix III Cost Data Line Costs Conductor Size Cost Basis (161kV Rural) Line Length Total Cost of Line Dove $106,000 15.57 miles $1,650,420 Drake $115,000 27.25 miles $3,133,750 $4,784,170 Capacitor/Inductor Bank Costs Bus Installation Cost Capacity Cost Basis Bank Size Total Bank Cost 5 $60,000 $300/100kVAR 100,000 kvar $360,000 5 $60,000 $300/100kVAR 100,000 kvar $360,000 10 $60,000 $300/100kVAR 2,500 kvar $67,500 10 $60,000 $300/100kVAR 15,000 kvar $105,000 13 $60,000 $300/100kVAR 15,000 kvar $105,000 16* $60,000 $300/100kVAR 10,000 kvar $90,000 17* $60,000 $300/100kVAR 10,000 kvar $90,000 $1,177,500 *Indicative of inductor banks.
Appendix IV One-line Diagrams Assignment I
Assignment II
Assignment III
Appendix V Assignment I Results Basic Plan
Line 5-11 Down