Knowledge of load characteristics in distribution feeders is critical to operating distribution networks. Do the loads include electric heating, air

Transcription

1 Knowledge of load characteristics in distribution feeders is critical to operating distribution networks. Do the loads include electric heating, air conditioning, or induction motors? This presentation shows how to use voltage regulators and their controls to estimate load parameters, which can help users make smart decisions to improve the operation of the distribution network. 1

2 Many utilities and grids employ intentional voltage reduction in order to reduce load during times of generation and/or transmission shortages. Voltage reduction can be controlled with tap changing transformers and distribution voltage regulators. On August 24, 2011, the Electric Reliability Council of Texas (ERCOT ) initiated Energy Emergency Alert Level 1 due to generation reserves dropping below 2,300 MW during a time of record high temperatures and electricity usage. As reserves continued to fall below 1,750 MW, ERCOT then initiated Level 2. According to the Operating Procedure Manual at the ERCOT Transmission and Security Desk, grid operators take the following actions at Level 2: Drop interruptible loads (load resources under contract). Begin blocking load transfers to other grids. Reduce customer loads by using distribution voltage reduction measures, if deemed beneficial. If the load on a feeder or bus is mainly resistive or constant impedance, then voltage reduction can be effective in reducing the current and power required by the load during times of power system duress. 2

3 Some traditional loads like light bulbs and cathode ray tube televisions are becoming extinct, while new loads like plug-in electric vehicles, liquid crystal display (LCD) screens, and compact fluorescent light bulbs are growing. This load diversity adds complexity to load parameter estimation. 3

4 The equation shown on the slide is used to represent the sensitivity of the real power requirements of a load for small voltage changes. In this equation, P is the new real power and P 0 is the previous real power (the difference between P and P 0 is the change in real power). V is the new voltage, and V 0 is the previous voltage (the difference between V and V 0 is the change in voltage). The exponent, K p, provides a measure of sensitivity of the load to voltage changes. It answers the question, for a given change in voltage, how do we expect the real power to change? The equation for reactive power, Q, is similar. These models are valid for voltage changes of plus or minus 10 percent. Assuming a constant frequency, the real power equation simplifies to that shown on the slide. If K p is 0, then the load is constant real power (e.g., induction motors). If K p is 1, then the load is constant current (e.g., fluorescent lighting). If K p is 2, then the load is constant impedance (e.g., an electric range or stovetop). 4

5 K p is also approximately the ratio of the power change in per unit (pu) to the voltage change in pu. This approximation provides a better idea of the value of K p. 5

6 To quantify the real power change with respect to voltage changes, we tested three loads: an incandescent light bulb, a compact fluorescent light bulb, and an LCD television. We connected the loads in parallel and monitored the currents independently. We varied the voltage from 114 to 120 V and from 120 to 126 V. On the slide, the yellow line is the power consumed by the incandescent light bulb. As shown by the plots, power increases when voltage increases. Using the equation to estimate load parameters, K p has a value of

7 This slide shows the power consumed by the compact fluorescent light bulb. Notice that the power does not change as much as in the incandescent light bulb example. For this case, the K p value is

8 This slide shows the power consumed by the LCD television. In this case, the power barely changes when the voltage changes. The K p value is close to 0. 8

9 Now, let us look at a real power system. Voltage regulators and their controls are ideal for monitoring distribution loads. Power and voltage changes can be captured every time the regulator changes taps and are used to calculate K p. In the substation shown on the slide, the voltage regulators are located right after the feeder breaker. 9

10 Use the voltage and current measurements before and after the tap change to calculate the corresponding powers and K p coefficient for each tap change event. Typically, a voltage regulator takes less than 0.3 seconds to change one tap position. As shown on the slide, the voltage and real power change for a voltage regulator tap-up operation. 10

11 A system used to monitor six feeders in one distribution substation in the northwest region of the United States is shown on the slide. The system consists of the following: Voltage regulator Voltage regulator control Satellite-synchronized clock Rugged computer The control devices send voltage, current, and motor command information with the corresponding time stamp to the rugged computer 60 times per second. The rugged computer archives these data to monitor the loads 24/7. 11

12 This slide shows three weeks of active power for one of the feeders that supplies energy to a university. Observe the weekly pattern, which shows less power consumption during Saturdays and Sundays. 12

13 This slide shows the active power of a typical residential feeder during a 24-hour period. The lowest consumption happens at 6 a.m. The load has two peaks: one before lunch at 10 a.m. and the highest at 9 p.m. when everyone is likely at home. 13

14 The university feeder has a different pattern over a 24-hour period. The lowest consumption happens at 6 a.m., but the loading of the feeder does not have the double-peak pattern of a typical residential load. The loading at the university stays above 1.4 MW from 10 a.m. to 10 p.m. 14

15 The actual measurements of voltage and power of one the substation feeders before, during, and after the tap operation are shown on the slide. The green line shows the motor raise command. This command acts as a reference to obtain pre-tap and post-tap data for the K p estimation. K p is 1.8 for this tap-up operation. We performed K p calculations for each tap operation over 47 days. These data are shown on the following slides. 15

16 The K p value of a feeder with residential and rural loads exhibits normal distribution around 1.8, with a standard deviation equal to 0.5, for 432 tap operations. 16

17 On the other hand, the K p value of a feeder with residential and industrial loads does not have normal distribution. The slide shows the K p values for 221 tap operations. These loads do not have a predictable pattern. We can use the K p information about a feeder to make informed decisions when operating the power system. 17

18 Using load estimation results, we can reduce power consumption in loads with K p values greater than 1. With the knowledge of when load peaks happen, we can activate conservation voltage reduction for peak reduction. 18

19 Greg Hataway from PowerSouth Energy Cooperative worked with SEL to explore the benefits of conservation voltage reduction at the utility and has offered to share his experiences. 19

20 20