SUSTAINABLE FUTURE ENERGY 2012 and 10th SEE FORUM
Innovation for Sustainable and Secure Energy
21-23 November 2012, Brunei Darussalam
Decentralized Power Factor Correction
J. Hazra1, Balakrishnan Narayanaswamy1, Kaushik Das1, Ashok Pon Kumar1, Deva P Seetharam1, De Silva Liyanage2 and Sathyajith Mathew2
1IBM Research India, Bangalore
2UBD | IBM Centre, University Brunei Darussalam, Brunei Darussalam
Tel: (91) 9739257776, E-mail: .
Abstract: Power Factor (PF) is a measure of electrical efficiency and is given by the ratio of KW (kilo watt) to KVA (kilo volt ampere), where KW is actual power consumed by the load whereas KVA is total power delivered to the load. The power consumed is called active power and the remaining (KVA-KW) is called re-active power. Even though reactive power doesn’t do any actual work, it still needs to be generated and carried. When power factor is poor (pf<1), reactive power travels through the wires between the load and the utility grid, passing back and forth through the residential meter which makes the system less efficient by increasing the system loses. This also increases the monthly electricity bill when reactive power is charged. In order to improve the power factor, we propose a system and method that achieves power factor correction with minimal extra hardware through intelligent scheduling of electrical loads.
Keywords: Power system modeling, power flow study, power loss minimization.
1. INTRODUCTION
Improving energy efficiency is one of the key goals of smart grid initiatives across the globe. In practice, a significant portion (5-10%) of generated energy is lost in transmission and distribution (T&D) system. In some countries like India T&D loss is as high as 26%[1]. Hence, there is immense potential to improve the energy efficiency by minimizing the transmission and distribution loss in the grid. For example, according to Federal Energy Regulatory Commission (FERC), making the US grid even 5 percent more efficient would save 42 gigawatts of electricity, an amount equal to 42 coal power plants [2]. There are several factors for having such high T&D loss. Some these key factors are inadequate investment in T&D section which resulted in often overloading in distribution system, unplanned growth in sub-transmission and distribution for fulfilling the short term objectives, too many stages of transformation, improper load management, inadequate reactive power compensation, poor power factor appliances, etc [1]. Among these factors low power factor is recognized as a major concern in improving energy efficiency and power industry engineers are striving actively for a commercial and engineering method for its correction [3]. Power factor correction not only reduces system loses but also releases system capacity and improves voltage regulation which enables the utility to provide cheaper and easier services of the quality desired in modern industry[4][5]. Until recently, reactive power was mostly compensated by placing centralized capacitor banks near the load centers. However, there are several issues associated with capacitor banks. Primarily such solution is costly (typically US$100=KVAR), requires large space to install, and can only supply reactive power but they cannot absorb it. When faced with rapidly increasing load and voltage drop, capacitors become increasingly less effective and can actually contribute to the downward spiral characteristic of voltage collapse. For static VAR compensators, their reactive power output varies with the square of the voltage. This characteristic makes them poor at coping with voltage instabilities and preventing voltage collapses. Sometimes they explode due to sudden short circuit. Proper containment, fusing, and preventive maintenance can help to minimize these hazards, which can be expensive.
Power factor correction (PFC) schemes [3] have been proposed for quite some time in the field of power electronics. The main aim of this PFC has been to reduce Total Harmonic Distortion (THD) and thereby improving the power factor. Recently, there is a growing trend for providing incentive for good power factor and/or penalty for poor power factor. For example; Tokyo Electric Power Co. gives its retail customers in Japan, a financial incentive to improve their power factors through discounts on the base rate [7]. European Union implemented a directive, EN61000-3-2, which controls the harmonic content and power factor of many products that are sold to European countries [8]. Similarly power factor correction can also be achieved using other equipment like battery energy storage system (BESS) [6]. Hence, to avoid penalty, there is a growing trend to install power factor correction devices (like capacitors) at home or commercial places. Unfortunately, these capacitor banks are not only expensive but also compromise with power quality by injecting additional harmonics to the grid through frequent switching. All these methods require additional equipment for PFC. Further, mostly these methods concentrate on improving power factor of a certain load. However, there can be some loads which are inductive while some other loads are capacitive. If these inductive and capacitive loads are scheduled properly, it can improve the power factor consequently improving the grid efficiency. Another advantage
In this paper we propose a framework for improving the power factor at a residence, group of residences or commercial location by intelligent load scheduling without using any extra hardware. The basic idea of the reactive power compensation here is to prevent the reactive component of the current from flowing along the transmission elements. i.e. lines, cables and transformers, simply by co-scheduling appliances of opposite reactance polarity. The flow of these reactive currents will then be mainly confined within the loads. As a result, the transmission elements can be chosen to transmit only the active current component. Accordingly, the required sizes of these elements will be smaller. We propose to schedule the appliances following the customer specific preferences and needs. For example, scheduling inductive load (such as washing machines, dryers etc.) at the same time as capacitive loads (such as LCD tv) will improve the overall power factor. Proposed scheme is simulated and experimentally verified on Indian appliances and effectiveness of the method is illustrated in this paper.
2. POWER FACTOR
Power Factor (PF) is the difference between the power needed to perform work and the electrical energy needed so the work can be performed. The power factor of an AC electrical power system is defined as the ratio of the real power flowing to the load to the apparent power in the circuit, and is a dimensionless number between 0 and 1. Real power is the capacity of the circuit for performing work in a particular time. Apparent power / total power (S)in kVA is the complex sum of real power / active power / true power (R) in kW and reactive power (Q) in kVAr. It can be observed from the power triangle below that the power factor (p.f.) is given by the cosine of the angle between total power and the true power.
S = P + jQ
Power Factor = P / S = P / √(P2 + Q2)
Figure 1: Power factor triangle
If the load is inductive or capacitive the power consumed by this load is utilized to store energy instead of doing work. This power is called reactive power. An inductive load consumes reactive power whereas; a capacitive load generates reactive power. Even though reactive power doesn’t do any actual work, it still needs to be generated and carried. When power factor is poor, reactive power travels through the wires between the load and the utility grid, passing back and forth through the residential meter which makes the system less efficient by increasing the system loses. This also increases the monthly electricity bill when reactive power is charged.
3. PROPOSED METHOD
In this paper, power factor of any residence or commercial place has been improved through intelligent scheduling of available appliances. The main idea behind this is to restrict the back and forth movement of reactive power between grid and residence/commercial place. Here appliances are scheduled in such a way that reactive power travels within appliances which drastically improves power factor, reduces network loss, relives line capacity and differs T&D investment. Appliances are classified into three groups i.e. resistive, inductive and capacitive. Inductive and capacitive appliances are tried to schedule together so that reactive power circulates between inductive and capacitive appliances. This is illustrated using the following picture. When either inductive or capacitive loads are scheduled, reactive power circulates between main grid and appliances as shown in Figure 2. On the other hand, when both inductive and capacitive loads are scheduled together, reactive power circulates within appliances. This improves the system efficiency significantly. However, there are several issues with rescheduling appliances. For example appliance like television can’t be rescheduled. Therefore, we introduced a inconvenience factor while scheduling the appliances.
Figure 2: Reactive power movement in a power network.
Mathematically the problem is formulated as follows:
;
Where
- E is a cost for consuming reactive power
- y is the level of reactive power consumption which is a function the power factor
- R is the rate for reactive power, d’ is number of appliances, is weight factor
- I1 and I2 are measures of the user inconvenience when the operating level and operating time (respectively) of the device are modified to improve power factor.
4. PROPOSED ARCHITECTURE FOR POWER FACTOR CORRECTION
The proposed architecture consists of three components i.e. smart meter, consumption profiler and scheduler as shown in Figure 3. Each of these components is described below:
4.1 Smart Meter
Smart meters can be installed at device level or mains level or both. These are already being installed at many homes due to smart grid initiatives. Smart meters measure active and reactive power consumption of appliances. From the cumulative signature of power consumption, appliance level active and reactive power consumption will be determined. This information will be used to classify the appliances (as resistive, inductive or capacitive) and to monitor the customer behaviour and comfort level.
4.2 Consumption profiler
It analyzes the data collected from smart meters. It reconstructs active and reactive power consumption of each appliance. The system can either utilize plug level meters (where available) or Non-intrusive. Appliance Load Metering (NIALM) along with a database of device characteristics toreconstruct appliance consumption profiles.
4.2 Scheduler
Scheduler can be installed at home, industry or commercial locations. It stores and processed consumption history and appliance profiles and computes schedules for appliances to improve power factor while respecting external constraints e.g. user availability, comfort, etc.
The detailed operation modularity of power factor correction scheme is presented in Figure 4.
Figure 3: Power factor correction architecture
Figure 4: Operation modularity of Power Factor Correction scheme
Table 1Measured capacity and power factor of typical appliances
Usage(hrs/day) / KW / PF
Television (Capacitive) / 8 / 0.2 / 0.70
Washing machine (Inductive) / 2 / 1.5 / 0.57
Water pump (Inductive) / 1 / 0.5 / 0.40
Freezer (Inductive) / 13 / 0.4 / 0.55
LED lights (Capacitive) / 5 / 0.2 / 0.70
Electric vehicle (Capacitive) / 7 / 2.0 / 0.75
Laptop charging (Capacitive) / 4 / 0.4 / 0.80
Dishwasher (Inductive) / 5 / 1.5 / 0.70
5. RESULTS
Power factor correction through intelligent load scheduling was simulated for a typical household. It was assumed that the household has appliances like television, washing machine, water pump, freezer, LED lights, electric vehicle and dishwasher. Typical power consumption profile of each appliance, collected using plug level meters, is shown in Table 1.
Figure 5: Power factor minimizing power schedule
Given the consumption characteristics in Table 1, we schedule the devices to minimize inconvenience and power factor as in Equation (1). For this we use IBMs ILOG CPLEX software [9]. The results of this scheduling are shown in Figure 5.
For the purposes of illustration, consider the 8:00AM time slot. Without scheduling the water pump and the freezer, the net reactive power component would be -1.64. However, after scheduling them, the net power factor drops to -1.22, a 25% reduction. This substantial reduction results in reduced generation, transmission and capex costs as discussed earlier. Similar improvements can be seen in other time slots, particularly in the evening time slots where peak consumption occurs.
6. CONCLUSION & FURTHER RESEARCH
This paper proposes a power factor correction scheme through intelligent load scheduling. Proposed approach improves power factor while considering customer or user related factors such as user preferences for device usage time, duration and delay, user preference for device usage intensity and change in intensity, user preference for device co-scheduling and re-scheduling. Improving power factor by intelligent load scheduling can reduce the reactive power requirement from grid and can improve system efficiency and reduce monthly electricity bill. As the proposed scheme avoids use of additional costly hardware like capacitor/inductor bank, it is highly cost effective and has a significant potential for real world implementation.
8. REFERENCES
[1] K. K. Kapil, “Reduction in transmission and distribution loses, an opportunity for earning carbon credits”, Available online:
Brandon Lorenz, “Smart Grid Addresses Energy Efficiency, Power Quality and Reliability Issues”, Available online: L. W. W. Morrow, “Power-factor correction,” Transactions of the American Institute of Electrical Engineers, vol. XLIV, pp. 1–7, Jan 1925.
[4] Y. Jiang, F.C. Lee, G. Hua and W. Tang, “A novel single-phase power factor correction scheme,” Eighth Annual Applied Power Electronics Conference and Exposition, pp: 287-292, 1993
[5] S. Basu and M.H.J. Bollen, “A Novel Common Power Factor Correction Scheme for Homes and Offices,” IEEE Transactions on Power Delivery, Volume: 20, Issue: 3, pp: 2257 - 2263, 2005
[6] D.K. Maly and K.S. Kwan, “Optimal battery energy storage system (BESS) charge scheduling with dynamic programming,” IEE Proceedings in Science, Measurement and Technology, pp: 453- 458, 1995
[7] Papalexopoulos, Alex D., and George A. Angelidis. "Reactive power management and pricing in the California market." Electrotechnical Conference, 2006. MELECON 2006. IEEE Mediterranean. IEEE, 2006.
[8] Harmonic Current Emissions, Guidelines tothe standard EN 61000-3-2, Available online at ILOG CPLEX Optimizer”, url :