Energy-Saving in Constant Speed Running Motors by Means of Adaptive Sinusoidal Voltage Shimon Limor R&D dept. PowerSines LTD. Or-Yehuda, Israel e-mail:
[email protected] Doron Shmilovitz School of Electrical Engineering Tel-Aviv University Tel Aviv, Israel e-mail:
[email protected] Abstract—A transformer-based, purely sinusoidal, energy- saving controller for induction motors is presented. The controller applies a three-phase transformer in a reconfigurable scheme to create three different voltages. A simple control scheme is then applied to set the 'right voltage' to supply the motor. Significant reductions in the real and reactive power consumption were noticed, especially in motors operated under low load. I. INTRODUCTION Energy-saving methods have always been an attention- grabber. At first, most of the interest was driven by the potential reduction in the running costs of electrical equipment. In recent years, concerns over the environment and CO2 emission, along with oil shortages and the ever- increasing price of electricity, have stimulated the quest for electricity-saving technologies. New standards are being enforced that set minimum limits on equipment efficiency. Utilities and industrial consumers are required to reduce polluting emissions and fossil fuel consumption, and sustainable energy sources are encouraged by most governments. Consequently, the 2000s have seen an increasing number of energy-saving devices and apparatuses for the operation of significant electrical loads, such as lighting and refrigeration. Because AC induction motors represent a dominant share of electricity consumption [1], increasing their efficiency is of high interest [2], [3]. In practice, it often happens that motors operate under low load. For instance, drive system considerations, such as start torque requirements, may result in a motor's rating being 100% higher than the rating required in normal running. In some systems, the motors are required to run at a pretty constant speed, but their load torque varies in a wide range. For example, escalators might be loaded by a single person (nearly no-load conditions) or fully loaded. Even worse, in this application, the drive system needs to be able to start under full load, which may result in a motor selection with a rating twice as high as the maximal nominal load. Furthermore, typical operating conditions may be at a load torque less than 30%. Although induction motors normally exhibit high efficiency (typically 85%-94%, depending on the motor's size and type) and a relatively high power factor when fully loaded, both of these factors deteriorate significantly if they are operated under a light load. II. MOTOR OPERATION UNDER REDUCED SUPPLY VOLTAGE Induction motors are designed for maximal performance in the region of nominal operation. Typically, they exhibit high efficiency (85%-94%, depending on the motor's size and type) and relatively high power factor when fully loaded [4], [5]. However, in practice, there are many situations in which motors operate under low load. For instance, motor ratings derived from starting requirements may be significantly higher than those required in normal running. Other examples include escalators, which are required to run at relatively constant speed even though they might be loaded by just 5% of nominal load (corresponding to a single person using the escalator). Moreover, they may be required to start at full load (for instance, when recovering from an interruption). It is well known that a motor's efficiency and power factor drop significantly when operated under low load torque, far from its nominal operation point [4], [5], see Fig. 1. This characteristic behavior can be understood from the motor equivalent scheme shown in Fig. 2; the iron losses, ΔPFe, are constant with respect to loading, which is the main reason for the decrease in efficiency at low load: 2 2 ' 2 ' 2 '2 ' 2 '2 ' 2 1 1 2 2 13 13 3 3 mech mech fe sI R P s sI R P P I R I R s η −⎛ ⎞ − Δ⎜ ⎟ ⎝ ⎠= −⎛ ⎞ − Δ + Δ + ⋅ ⋅ + ⋅ ⋅⎜ ⎟ ⎝ ⎠ (1) This research was supported by PowerSines LTD., Or-Yehuda, Israel. 978-1-4577-0541-0/11/$26.00 ©2011 IEEE 1083 Figure. 1. Efficiency and Power-Factor degradation with load reduction, reproduced from [1]. Figure 2. AC machine equivalent scheme. The main reason for power factor degradation at low load can also be seen from Fig. 2; the reactive power associated with the magnetizing current, ΔQx, is independent of load (and I2). Therefore, as the load (and thus the output power P2) decreases and ΔQx does not change, the power factor decreases. III. THE PURELY SINUSOIDAL CONTROLLER In order to attain a controllable voltage reduction, we apply a transformer in a reconfigurable scheme, as depicted in Fig. 3, in a manner that resembles a Y/Δ start. However, while a standard Y/Δ start facilitates only one voltage reduction step, phase-Y phase-Δ 1 3 V V= , the proposed controller offers three reduced voltage levels. The transformer’s secondary windings are connected in series with the line voltage; thus the Figure 3. The 3 phase reconfigurable-autotransformer based controller. transformer operates as an auto transformer, reducing its required rating (The advantages of autotransformer topology have also been extended to electronic power processing circuits ). The primary windings connect to the mains through four contactors which enable variability of the connection. Three reduced voltages are generated by different combinations of the contactors K1-K4, two at a time. The different voltage vector relations resulting in motor reduced voltages are illustrated in Fig. 4, where the secondary windings voltage is marked by the brown color. This voltage might take the value of either llk V⋅ or phasek V⋅ , depending on the primary windings configuration (where k is the transformer's phase to phase transfer ratio). μ μ μ a b c a' b' c' 1084 (a) (b) (c) Figure 4. Three reduced voltage combinations Using Fig. 4, the motor line voltages, ll motorV − , can be easily formulated yielding (2), (3), and (4). Step 1: K2 and K3 on (while K1 and K4 are off): With K2 on, a virtual neutral point is formed and the primary becomes Y connected. However, it should be noted that with K3 on, the primary Y connection is rotated by 1200 with respect to the line voltages (Vl2 is applied to the A winding, Vl3 to the B winding, and Vl1 to the C winding). The secondary reflected voltages connect in series with the line voltages. Denoting the phase-to-phase winding ratio by a, the line voltage supplied to the motor is calculated: 21ll motor lineV k k V− = − + ⋅ . (2) Step 2: K1 and K4 on (while K2 and K3 are off): With K4 on, the primary is Y connected as well. K1 on sets this Y connection to be in the normal phase relation with respect to the line voltages (Vl1 is applied to the A winding, Vl2 to the B winding, and Vl3 to the C winding). The construction of the motor side voltages yields: ( )1ll motor lineV k V− = − ⋅ . (3) Step 3: K1 and K3 on (while K2 and K4 are off): In this case, the primary becomes Δ connected, resulting in a secondary reflected voltage higher by a factor of 3 compared to the previous cases, yielding the motor side voltage: 21 3 3ll motor lineV k k V− = − + ⋅ (4) By trading off the transformer rating, the typical range of motor loading, and the range of effective voltages that may be applied to motors without the risk of stalling, we have chosen the lowest voltage level to be 220V line to line. The phase to phase turn's ratio is 0.275. Thus in the configurations in which the primary sees the line voltage (400V), 110V develop across the secondary. This results in the following line voltages that can be applied to the motor: 400V (no voltage reduction), 358V, 290V, and 253V (these are no-load voltages, not accounting to the transformer's voltage drop). IV. EXPERIMENTAL A fractional horse power motor was used to demonstrate the controller operation. The motor's nameplate: P=0.37kW, 400/230VRMS, cos(φ)=0.76 and n=1390rpm. The motor was driving an air blower. When the motor was supplied by the nominal voltage of 400V it run at n=1450 rpm and consumed: P=365Watt, Q=575VAR and S=681VA. Thus the resulting motor's PF was cos(φ)=0.54. When the controller was feeding the motor, it supplied the motor with a voltage of 290V (theoretically) and the measured data was: n=1390rpm, Vll=285V, P=308Watt, Q=417VAR and S=518VA, thus cos(φ)=0.6. Comparing the two situations one can note a 15% reduction of real power, and a 27% reduction of reactive power consumption. The speed decreased only by 4%, which in a sense justifies our assumption of nearly constant speed load (in any case a load that requires no speed control). llk V⋅ a a a a' a' a' b b' b b b' b' c' c c' c c' c 3 llk V⋅ 3 llk V⋅ 1085 Assuming a quadratic load characteristics, a speed drop=4% implies an output power reduction of Δpout~8%. Thus, the direct power saving (only due to reduced losses), are approximated as: %Saving=%Δpin-%Δpout=15%-8%=7%. It should be noted that this is only the directed saving in terms of motor's real power consumption and there are additional savings in the system due to current reduction. Figure 5 shows the controller's input (upper traces) and output (lower traces) voltage and current. Owing to the light load, the line voltage (orange) is reduced from 400VRMS to 220VRMS (obviously, the current increases accordingly). These waveforms were obtained with no filter whatsoever. The pure sinusoidal waveforms should be noted, particularly the current drawn from the mains (blue trace), which allows utilization of PFC capacitors, and the voltage applied to the motor (purple trace), which induces no torque reduction as switching-type controllers would. Experimental results show a 5% to 15% reduction in real power losses and an up to 50% reduction of the reactive power consumption. Furthermore, no harmonic currents are generated as would be the case with phase-controlled rectifiers or PWM ASDs. This controller is also suitable for soft start (within an appropriate control regime), exhibiting the same benefits: no harmonic pollution, fine operation in the presence of power factor correction capacitors and extremely high reliability. Figure 5. Input and output waveforms V. DISCUSSION In addition to electricity savings (i.e., cost savings of typically 5% to 15% through direct and indirect mechanisms), the controller provides a certain degree of voltage stabilization, surge protection, and harmonics mitigation (due to the series choke present in the transformer connection). From the point of view of the electrical power utility, deployment of such controllers would reduce peak demand and thus the power system stress and likelihood of instability. Furthermore, the pressure to increase generation is associated with peak power demand. This is a heavy economic burden, which is relaxed due to peak demand reduction attained by utilizing the proposed controller. Furthermore, harmonics generation is reduced and immunity to the presence of harmonics is improved. At present, over 50% of the load consists of modern electronic non linear loads that generate harmonics [1]. There are many detrimental effects due to harmonics, such as the blasting of power factor correction capacitors, protection devices tripping, and the need for bulky expensive filters. The transformer-based controller has a double impact in this regard; it does not generate any harmonics and it filters out harmonics generated by other sources. REFERENCES [1] D. Shmilovitz, Duan, J., D. Czarkowski, Z. Zabar, and S. Lee, “Characteristics of modern nonlinear loads and their influence on systems with distributed generation”, Int.J. Energy Technology and Policy, 2007, Vol. 5, No. 2, pp.219–240. [2] B. Sutopo, S.F.D. Widjaya, "Energy saving algorithm on induction motors controlled by a 68HC11 microcontroller system using fuzzy logic approaching," IEEE Proceedings. 4th International Conference on Power Electronics and Drive Systems 2001, vol. 1, pp. 59 – 61. [3] C.D. Pitis, , M.W. Zeller, "Power savings obtained from supply voltage variation on squirrel cage induction motors," IEEE Electric Power and Energy Conference, Canada 2008. pp. 1 – 3. [4] Mulukutla S. Sarma, Electric Machines Steady-State Theory and Dynamic Performance , second edition, PWS Publishing Company, 1994. [5] A. E. Fitzgerald, C. Kingsley Jr., S. D. Umans, Electric Machinery. Fifth (& forth) Editions, McGraw-Hill 1990. [6] D. Shmilovitz, “Time variable transformers operating at near unity transfer ratio and some possible applications”, IEE Trans. on Power applications, vol. 151, issue 2, March 2004, pp. 161-168. [7] T.A. Lipo, "The Analysis of Induction Motors with Voltage Control by Symmetrically Triggered Thyristors," IEEE Transactions on Power Apparatus and Systems, vol. PAS-90, no. 2, 1971 , pp. 515 – 525. [8] L. Holmes, "Energy-saving motor controllers - background and update," Electronics and Power vol. 28 , no. 3, 1982. pp. 232 – 235. [9] Farr, L.B., Farr, T.A, "Considerations in Medium Voltage Reduced Voltage Motor Starting the Good, the Bad and the Ugly," Pulp and Paper Industry Technical Conference, 2007. Conference Record of Annual 2007 , pp. 220 – 226. 1086