UNSW - SCHOOL OF ELECTRICAL ENGINEERING AND TELECOMMUNICATIONS TELECOMMUNICATIONS ELEC4612 POWER SYSTEM ANALYSIS LABORATORY EXPERIMENT 4 ECONOMIC DISPATCH AND OPTIMAL POWER FLOW 1. AIMS: With the aid of the PowerWorld Simulator program, the objectives are: To analyse the problem of economic dispatch. To analyse the problem optimal power flow. 2. BACKGROUND: 2.1 Economic Dispatch: Economic dispatch is a method that determines the allocation of the loads to the electricity generation resources in the power system, and minimizes the generation cost. Consider an ideal system with n generating units, neglecting unit constraints and system losses. The operating cost of generator i is C , which varies with the real power output P of the i i generator, and (1) Ci a bPi cPi 2 dPi 3 fu fuel el cost ost $/hr hr $/ where a, b, c, d are the coefficients of the cost curve. Generally, “a” represents the fixed costs that do not vary with the generator output, such as the cost of installing the generators; “b”, “c”, “d” are coefficients related to generator power outputs. The fuel input to the generator is in Btu/hr and represented by bPi cPi 2 dPi 3 . The fuel cost is in $/Btu. The operating cost can be reduced by optimizing the operation strategies. Practically, the curve of operating cost C i versus generator output is a piecewise-continuous and monotonically increasing curve. The discontinuities may be caused by the incremental firing of the equipment, like additional boilers or condensers. The incremental cost curve, dCi dPi versus Pi , which indicates the cost of producing one more MW power by the electricity generating unit, can be obtained by taking the slope or derivative of the operating cost curve. When C i consists of only fuel costs, dCi dPi specifies the heat rate of the unit. Heat rate is the ratio of the amount of heat energy required in Btu to generate one more MW of power, which indicates the fuel efficiency of the generator over its operating range. The lower this number is, the higher the efficiency of the unit is. Generally, generators reach their highest efficiency somewhere in the middle of their operating range, and become least efficient at the minimum and maximum MW output. For an area consists of n units which operate on economic dispatch, the total operating cost of the area is CT n C i C1 P1 C 2 P2 C n Pn (2) $/hr i 1 Neglecting system syste m losses, the total load demand d emand in the area is Elec4612 - Experiment 4 – Economic dispatch and optimal power flow Page 1 PT n P P P 1 i 2 Pn (3) i 1 Assuming the load demand PT is constant, the economic dispatch problem becomes the determination of the values P1 , P2 , Pn that minimize the total area operating cost CT , and the sum of unit outputs should be equal to the total load demand. The minimum total operating cost occurs when CT CT CT T 1 2 dC P dP P dP P dPn 0 1 2 n From Equation 2, the above can be written as: dC n dC1 dC 2 dCT dP1 dP2 dPn 0 dP1 dP2 dPn (4) (5) Since the load demand PT is assumed constant, dP1 dP2 dPn 0 (6) Multiplying Equation 6 by and subtracting the resulting equation from Equation 5, the following equation can be obtained, dC dC1 dP1 2 dP2 dP2 dP1 dC n dPn 0 dPn (7) According to Equation 6, Equation 7 is satisfied when dC1 dC 2 dP dP 2 1 dC n 0 dPn (8) Thus, dC1 dP1 dC 2 dP2 dC n (9) dPn Therefore, all units in the system should operate at the same incremental operating cost under economic dispatch, which is a criterion for the solution to economic dispatch problem. dC1 dC 2 dC n (10) dP1 dP2 dPn The criterion given in the Equation 10 can be explained as follows: neglecting the system loss, if a unit operates at a higher incremental cost than others, the output power shift from this unit to the unit with lower incremental cost will certainly reduce the total operating cost in the system. Thus all the electricity generating units must operate at the same incremental cost to achieve the minimum operating cost. If the unit’s output constraint, Pi min Pi Pi max , is considered, the units that have reached their limit values are held at their limits, and the others that are not at their limits will distribute remaining loads equally. In this case, the incremental operating cost of the area is determined by the units that are not at their limits. When the transmission losses are included, the problem becomes more complicated. The total cost of transmitting 1MW power includes the incremental cost and the cost due to the transmission losses. The unit with lower incremental operating cost may be so far away from the load centre that the total cost is higher than those with higher incremental cost. The total load demand is as follows: n PT Pi PL P1 P2 Pn PL (11) i 1 Elec4612 - Experiment 4 – Economic dispatch and optimal power flow Page 2 where PL is the total transmission loss in the system. Generally, PL depends on the unit outputs P1 , P2 , , Pn . Equation 6 can be rewritten as dP1 dP2 P P dPn L dP1 L dP2 P2 P1 PL dPn 0 Pn (12) Thus, Equation 7 can be rewritten as dC1 dC2 PL PL d P 1 dP2 d P P d P P 1 2 1 2 According to Equation 12, Equation 13 is satisfied when PL PL dC1 dC dC2 dCn dP1 P1 dP2 P2 dC P n L dPn 0 Pn dPn dPn PL 0 Pn (13) (14) or dC1 dP1 dC 2 dP2 P 1 L dPn Pn dC n (15) Equation 15 is the new criterion when the transmission losses are considered. When the transmission losses are negligible, which means PL P1 0 , this equation reduces to Equation 9. 2.2 Optimal Power Flow: Optimal Power Flow (OPF) is a generalized non-linear programming formulation of the economic dispatch problem with system constraints. Under economic dispatch, only output constraints of electricity generating units are considered. However, each transmission line or transformer has a limit as specified by its power rating. The solution to the Optimal Power Flow problem not only solves the dispatch problem economically, but also follows the transmission network constraints. An OPF problem is typically a non-linear programming problem with multi-objective, which aims to optimize the steady-state performance of a power system network and satisfy the equality and inequality constraints at the same time. t ime. It can be formulated as: f x, u Minimize g x, u 0 (16) h x, u 0 subject to. where x is the vector of state variables of a power system network which include slack bus real power output PG1 , load bus voltage VL , generator reactive power outputs QG , and transmission line loadings S l . u is the vector of control variables, which are self-constrained and include generator voltages VG , generator real power outputs except at the slack bus, transformer tap settings T , and reactive powers QC injected from shunt VARs. Therefore, x and u can be expressed as: x PG1 , VL1 VLN , QG1 QGN , S l1 S ln (17) u VG1 VGN , PG2 PGN , T1 Tn , QC1 QC N OPF is the optimized power flow, and the OPF equality constraints g x can be expressed as: Elec4612 - Experiment 4 – Economic dispatch and optimal power flow Page 3 N PP PGi PDi Vi YijV j cos i j ij i N j1 (18) N QP QGi QDi Vi YijV j sin i j ij i N j1 where PGi , QGi — real and reactive power generation at bus i; PDi , Q Di — real and reactive power demand at bus i; Vi , i Yij , ij — voltage magnitude and angle at bus i; — magnitude and angle of bus admittance. The system operational and security limits which are represented by the OPF inequality constraints hx can be shown as follows: PGi min PGi PGi max (i 1, QGi min QGi QGi max Vi min Vi Vi max S li S li max (i 1, (i 1, (i 1, , N) , N) (19) , N) , n) There are many methods for solving OPF problem. The most popular traditional optimization techniques are linear programming, Newton-based techniques, sequential quadratic programming, and generalized reduced gradient method. However these techniques are naturally local optimizers and each technique only suits a specific OPF problem. Nowadays, various modern algorithms, including Genetic Algorithm (GA), Evolutionary Programming (EP), Artificial Neural Network (ANN), Simulated Annealing (SA), Ant Colony Optimization (ACO), and Particle Swarm Optimization (PSO), have been proposed to avoid local minima. They allow engineers to solve the OPF problem more efficiently and effectively for large power systems. Elec4612 - Experiment 4 – Economic dispatch and optimal power flow Page 4 3. SIMULATIONS: Procedure: This experiment involves the use of PowerWorld Simulator. A 10-bus power system is shown in Figure 1 below. Figure 1: A 10-bus power system. This system consists of 10 buses, and three generators (G1 at bus 1, G2 at bus 10, G3 at bus 3). The corresponding data of the system is listed in Table 1 and Table 2. AGC are available for all the generators and enforced MW limits are applied. Table 1: Bus data Bus No. P P P Q Generated max.(MW) Generated min.(MW) Load (MW) 1 300.0 80.0 2 0.0 3 Q Q Voltage Load (MVar) Bus Type* Generated max.(MVar) Generated min.(MVar) Level (kV) 0.0 0.0 2 50.0 -6.0 13.8 0.0 21.7 12.7 3 0.0 0.0 13.8 200.0 60.0 54.8 19.0 2 60.0 -6.0 13.8 4 0.0 0.0 77.3 16.6 3 0.0 0.0 13.8 5 0.0 0.0 0 0 3 0.0 0.0 13.8 6 0.0 0.0 0 0 3 0.0 0.0 13.8 7 0.0 0.0 9 5.8 3 0.0 0.0 13.8 8 0.0 0.0 6.1 1.6 3 0.0 0.0 13.8 9 0.0 0.0 57 5.8 3 0.0 0.0 13.8 10 150.0 30.0 14.9 5.0 1 80.0 -6.0 13.8 *Bus Type: (1) swing bus, (2) generator bus, (3) load bus. Elec4612 - Experiment 4 – Economic dispatch and optimal power flow Page 5 Table 2: Line data From Bus To Bus Resistance (p.u.) Reactance (p.u.) Line Charging (p.u.) 1 2 0.01938 0.05917 0.0528 1 5 0.05403 0.22304 0.0492 2 3 0.04699 0.19797 0.0438 2 4 0.05811 0.17632 0.0374 2 5 0.05695 0.17388 0.034 3 4 0.06701 0.17103 0.0346 4 5 0.01335 0.04211 0.0128 4 7 0.03181 0.08450 0.0 4 10 0.12711 0.27038 0.0 5 6 0.09498 0.1989 0.0 5 8 0.12291 0.25581 0.0 5 9 0.06615 0.13027 0.0 6 7 0.08205 0.19207 0.0 8 9 0.22092 0.19988 0.0 9 10 0.17093 0.34802 0.0 *Rating for transmission lines are 150MVA. The fuel cost is 1.00 $/Mbtu, and the operating costs of the generators, using quadratic cost model, can be specified as follows: C1 2 P1 0.375 P12 fuel cost $ / hr C2 1.75 P2 1.75 P2 fuel cost $ / hr C3 3.25 P3 0.834 P3 fuel cost $ / hr 2 2 where P1 , P2 , P3 are in MW units. 1. Obtain the Incremental the of Fuel curve ofunder the generators the same scale, and forecastCost the curve outputand trends theCost generators Economic using Dispatch state. 2. Run the simulation and write down the total MW losses, the total MVar losses, and the hourly cost of the system, under Economic Dispatch state. Record the marginal cost as well. To change the AGC Status, select Case Information Aggregation Areas, and then double-click on the AGC Status field. 3. Repeat (2) by using LP OPF algorithm. To solve the OPF, first toggle the AGC Status field to OPF. Then select Add Ons Primal LP to solve the power flow using the LP OPF algorithm. Comment on the results. 4. Add a capacitor bank to bus 9. Gradually increase the voltage of bus 9 above 0.9800, and resolve the OPF problem. Comment on the results. Remove the capacitor bank before proceeding to the next part. Elec4612 - Experiment 4 – Economic dispatch and optimal power flow Page 6 5. If the maximum MW output of G1 is 150MW, what is the marginal cost of the area in OPF? What if the maximum MW output of G1 is 100MW? Set the maximum MW output of G1 back to 300MW before proceeding to the next question. 6. Increase the fuel cost of G1 to 2.0 $/MBtu. Resolve the OPF problem. Write down the total MW losses, the total MVar losses, the hourly cost and marginal cost of the system. Comment on the results. 4. DISCUSSION: 1. Explain the difference between economic dispatch and optimal power flow. 2. Comment on the relation between the output of the generators and the marginal cost of the area. _____________ ____________ _ Elec4612 - Experiment 4 – Economic dispatch and optimal power flow Page 7