Comparison of the Performance of a PWM and a Binary Weighted Load Induction Generator Controller

Comparison of the Performance of a PWM and a Binary Weighted Load Induction Generator Controller

Available online at www.sciencedirect.com ScienceDirect Energy Procedia 93 (2016) 133 – 140 Africa-EU Renewable Energy Research and Innovation Sympo...

513KB Sizes 12 Downloads 532 Views

Available online at www.sciencedirect.com

ScienceDirect Energy Procedia 93 (2016) 133 – 140

Africa-EU Renewable Energy Research and Innovation Symposium, RERIS 2016, 8-10 March 2016, Tlemcen, Algeria

Comparison of the performance of a PWM and a binary weighted load induction generator controller Richard Sseruwagia*, Richard Okoua Ambrose Bukenyaa, George Sendifaa, Joanita Komugabea, Peter Kyeyunea a

Makerere University, Dept. Of Electrical and Computer Engineering,, P.O. Box 7062. Kampala Uganda.

Abstract In Uganda only 18.2 % of the population has access to electricity. Pico and micro hydro systems could be used to implement off grid energy generating systems to increase energy access. One of the key challenges is the high cost of electromechanical systems. The renewable energy business incubator in partnership with the Dept. of Electrical and Computer Engineering Makerere University pioneered three projects to develop electronic load controllers and induction generator controllers for pico and micro power schemes. The controllers developed would be used to reduce the costs of implementing off grid power schemes. Load controllers are used to control power quality without the need of a costly mechanical governor. The controllers developed were able to control the voltage and frequency of the alternators and motor generators. Total voltage harmonic distortion of 3% and 2.6% are exhibited by the PWM and Binary Induction generator controller. © by by Elsevier Ltd.Ltd. This is an open access article under the CC BY-NC-ND license © 2016 2016The TheAuthors. Authors.Published Published Elsevier (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of RERIS 2016. Peer-review under responsibility of the organizing committee of RERIS 2016 Keywords: PWM IGC; binary IGC; induction generator load controller; Renewable Energy Business Incubator.

* Corresponding author. Tel.: +256 774 595 294. E-mail address: [email protected]/[email protected]

1876-6102 © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of RERIS 2016 doi:10.1016/j.egypro.2016.07.161

134

Richard Sseruwagi et al. / Energy Procedia 93 (2016) 133 – 140

1.Introduction With only 18.2 % [1] of the population having access to electricity, pico and micro run of the river hydro power schemes are a promising solution to increase access to electricity in Uganda in regions with hydro power resources. It is estimated that Uganda has a 2000 MW hydro power potential, however only less than 10 % [2] has been developed. There has been increased private sector involvement in the energy sector due to the unbundling of the former Uganda Electricity Board (UEB). The Rural Electrification Agency was mandated to extend electricity transmission lines to the different parts of the country [3]. In some of these places it is still not economically viable to extend these lines. It is in such circumstances that off grid energy systems are recommended to stimulate the electrical load for a period of at least 5 years. One such case is the Bwindi hydro power scheme implemented by GIZ [4]. Cost is a key deterrent to the development of these micro hydro power schemes mainly due to the complex and costly control systems required [5]. The development of a locally fabricated load controller was perceived as one of the solutions to address this challenge. 2.Methods Two induction generator micro controller based controllers were developed. One was developed based on the pulse width modulation (PWM) technique, the other on the binary weighted load technique. These controllers were tested on a test bench comprising a 0.373 kW 220 Vac C-2C induction generator driven by a 1.1 kW 400 Vac 3 phase 50 Hz induction motor powered by a variable speed drive to vary the speed. Dump loads sized 300 W, 200 W, 100 W, 50 W and 25 W were used to obtain the different binary weighted load combination requirements. The wave forms as a result of each of the controllers were read using a power quality analyser (PQA) and studied. 3.Results and discussion 3.1.Total voltage harmonic distortion (Vthd) From results obtained with the power quality analyser (PQA), it is noted that the total voltage harmonic distortion averages to about 3 % for the PWM IGC when the controller is switched as long as no overload occurs. The Vthd value is low because the pulse width modulation is done at a relatively high frequency of 3.9 kHz and partly due to the action of the excitation capacitors that also act as filters. If, however the dump load is so oversized, the Vthd would increase tremendously and a specially designed harmonic filter would have to be incorporated. At overload, the Vthd is about 2.2 % as shown below in Fig. 1 and this is inherent to the system. 

Richard Sseruwagi et al. / Energy Procedia 93 (2016) 133 – 140

Fig. 1. Total harmonic distortion plot of the system under a PWM IGC.

The voltage waveform with the controller on at no user load is shown below in Fig. 2. The waveform is seen to have some distortion due to the PWM being done by the controller. 

Fig. 2. The output waveform of the system under a PWM controller.

With the Binary IGC the total voltage harmonic distortion is about 2.6 % on average. This is partly due to the inherent nature of the generator, the excitation and due to switching that occurs as the dump load combinations are changed as shown below in Fig. 3

135

136

Richard Sseruwagi et al. / Energy Procedia 93 (2016) 133 – 140

Fig. 3. Total harmonic distortion plot with a binary IGC.

The generator voltage waveform with the controller on and no user load is shown below in Fig. 4. It is observed that there is minimal distortion to the waveform resulting from the controller. This is expected of a binary controller. 

Fig. 4. The output wave form of the system under the Binary IGC.

3.2.Transient analysis Based on different loading scenarios, transient analysis of the two IGCs was carried out and the waveform obtained for the PWM and Binary IGCs are shown below in Fig. 5 and Fig. 6.

Richard Sseruwagi et al. / Energy Procedia 93 (2016) 133 – 140

Fig. 5. Transient analysis of the system under the PWM IGC.



Fig. 6. Transient analysis of the system under the binary IGC.

3.3.Detailed analysis 3.3.1.8 W CFL switched on Although the PWM method achieved success and voltage regulation, a few key points were noted such as when connected to the highly inductive loads (in this case compact fluorescent lamps), it is noted that some harmonics are introduced in the system. As seen on the waveform, there is a change in the smoothness of the waveform as shown below in Fig. 7;

137

138

Richard Sseruwagi et al. / Energy Procedia 93 (2016) 133 – 140

Fig. 7. PWM waveform with an 8 W CFL bulb.

On the other hand, using the binary weighted loads method, this is not a problem in the system. The waveform always maintains the smoothness even with compact fluorescent lamp loads. The waveform for the binary circuit is shown below in Fig. 8; 

Fig. 8. Binary IGC with an 8W CFL bulb.

The Compact Fluorescent Lamp is a switched mode power supply, which continually switches between states to limit power dissipation. The switching currents in the system cause electrical noise problems. These cause harmonics in the PWM system when interacted with the IGBT fast switching with this nature of loads.

Richard Sseruwagi et al. / Energy Procedia 93 (2016) 133 – 140

3.3.2.100 W Incandescent bulb With the incandescent 100 W bulb connected to the system, the PWM waveform is noted to have a momentary change in amplitude of the waveform for a duration of about 0.112 ms. The amplitude is seen to sharply decrease when the load is switched into play, but quickly settles to a voltage within the tolerance range of ±5 % from the reference voltage (Fig. 9). This settling time is fast enough to avoid instability.

Fig. 9. PWM IGC waveform for 100 W incandescent bulbs.

With the binary circuit is connected to the system, the voltage dips at first, rises and slightly overshoots the set point, then stabilizes. The time to achieve stability is about 0.24 seconds, which is somewhat fast enough for the controller to track rapid changes in user load without resulting into instability as shown in Fig. 10. 

Fig. 10. Binary IGC waveform for 100 W incandescent bulb.

139

140

Richard Sseruwagi et al. / Energy Procedia 93 (2016) 133 – 140

In this scenario, we notice that the PWM system switches and settles at a slightly faster rate to a value within the tolerance range, whereas the binary weighted system switches at a slightly lower rate but settles back to the reference voltage. However, the settling time and response time are dependent parameters and can be altered by changing the gain values in the controller code. 4.Conclusions The main advantage of PWM is that it requires a simple electronic circuit for steering the switching device. The disadvantage is the high dissipation in the controller since the generator voltage first has to be rectified before it can go to the power transistor itself. Thus there is need for a large heat sink. Binary weighted loads have several disadvantages such as the number of dump loads each with its connections, wires and switching device. To achieve smooth regulation, these dump loads must all have exactly the right capacity. With a low number of dump loads, steps between dump load combinations remain too large and the system cannot regulate smoothly. In summary, both the controllers achieve both voltage control and regulation. Based on the availed results, the binary weighted system is a better method for implementing the induction generator controller. Even with the complexity and cost, it is not affected by the nature of the loads connected, the voltage settles back to its reference voltage with a variation in load and waveform distortion is not produced. No harmonic filter is required for binary controllers, but may be required by a purely PWM controller. References [1] UBoS, Uganda Housing Census Statistical Abstract, Uganda National Bureau of Statistics, 2014. [2] A.A.Asere and K.O.Adeyemi, "A REVIEW OF THE ENERGY SITUATION IN UGANDA," International Journal of Scientific and Research Publications,, 2014. [3] REA, Rural Electrification Strategy and Plan 2013-2022, Ministry of Energy and Mineral Development, (Rural Electrification Agency), 2013. [4] R. J. v. d. P. a. A. Kyezira, November 2015. [Online]. [5] N. L. U. Eva Maate Tusiime, November 2015. [Online]. Available: http://www.smallhydroworld.org/fileadmin/user_upload/pdf/Africa_Eastern/WSHPDR_2013_Uganda.pdf. [6] G. K. Bhim Singh, "An improved electronic load controller for an isolated asynchronous generator feeding 3-phase 4-wire loads," IETE J Res,, 2009. [7] P. M. a. N. Smith, PICO HYDRO FOR VILLAGE POWER: A Practical Manual for Schemes up to 5 kW in Hilly Areas. Edition 2.0, 2001 . [8] N. Smith, Motors as generators for micro-hydro power, 4th edition, Southampton Row: ITDG Publishing, 2001. [9] P. Maher, The pico power pack: a new design for pico hydro. [10] J. B. Ekanayake, "Induction generators for small hydro schemes," Power Engineering Journal, Vol. 16, No. 2, Apr. 2002. [11] R. J. a. B. S. S. S. Murthy, "A Practical load controller for standalone small hydro systems using self excited induction generator," IEEE Proc. of PEDES for Industrial Growth, Dec 1998. [12] N. L. U. Eva Maate Tusiime, November 2015. [Online]. Available: http://www.smallhydroworld.org/fileadmin/user_upload/pdf/Africa_Eastern/WSHPDR_2013_Uganda.pdf.