Abstractâ The use of domestic renewable energy, modern appliances and electronic gadgets at home are becoming more popular as technology develops.
Probabilistic Model for Residential Load Profiles and Power Quality Improvement G. Gupta and W. Fritz Cape Peninsula University of Technology, South Africa, Western Cape, Bellville, South Africa
Abstract— The use of domestic renewable energy, modern appliances and electronic gadgets at home are becoming more popular as technology develops. This paper investigates the electricity requirements of these appliances and their influence on power quality (PQ). PQ issues vary from device to device depending on applications in the residential, commercial and industrial sectors. PQ problems are very challenging in electrical distribution systems. A literature survey was done on existing models of residential loads affecting PQ. The residential loads investigated in this paper are domestically used photo voltaic (PV) roof top systems and small wind turbines. This paper further focuses on voltage unbalance due to rooftop PV in residential areas and factors which affect wind turbine flickers. Index Terms—Photo Voltaic, Power Quality, Wind Shear, Tower Shadow
1
INTRODUCTION
Energy saving, improved energy efficiency and environmental protection are ubiquitous topics in society. Despite many efforts to save energy, demand for electricity is expected to grow and grow much faster in comparison with other energy sources over the next few decades. Power Quality (PQ) has become one of the most prolific buzzwords in the power industry. It is basically the fitness of electrical power to consumer devices. Due to PQ problems like voltage sags, voltage swells, interruptions (micro and long), voltage unbalance, voltage fluctuation and harmonic distortion, electrical devices may not operate optimally. These PQ problems affect residential loads. The paper focuses first on the probabilistic model which directly estimates energy consumption by using extensive information on end use and end users such as appliances, the customer use, their age, size of houses etc[1]. Then, we discuss Roof top PhotoVoltaic (PV) systems and wind turbines which contribute to residential loads. Rooftop PV systems are basically those PV systems that has its electricity generating solar panels mounted on the rooftop of a residential or commercial buildings. PV modules, mounting systems, cables and solar inverter are its main components. Roof top PV systems on residential buildings typically have ratings and capacity of about 5 to 20 Kilowatts. These are semiconductor devices, which are capable of converting incident solar energy into dc current, with efficiencies varying from 3% to 31%, depending on the technology, temperature, design and the
material of the solar cell [2].Reverse power flow is a major problem affecting power quality of the voltage of electric power distribution when a growing number of large installations is large. So, we focussed on analysis of voltage unbalance at the planning stage [3]. Then the paper discusses the factors which affect flickers in wind turbines. Now a days, wind turbines play an important role in producing electrical energy among other distributed generation sources [4]. Voltage fluctuation causes flicker in lamps [5,6]. 2
PROBABILISTIC MODEL
This model estimates the electricity requirements of household appliances such as refrigerators, TV, light, kitchen devices etc, depending upon the customer’s use, age and size of houses in South Africa. As a survey, we decided to make a family type category with their distribution, which is indicated in Table 1[1]. TABLE I. Distribution of Family Category over samples Family Job % Type Full Time 5 Single Part Time 1 Retired 8 Full Time 8 Couple Part Time 0 Retired 30 2 full time + children 12 without retired member Family 1 full time + children 24 Family with retired 11 member
The survey in Table I was conducted in 2010. For data analysis, we have categorized household into 5 groups[1]. 1. Entertainment such as DVD player, Play Station, TV, etc. 2. Communication such as Mobiles, Cordless Phone, etc. 3. Office such as PC, Laptop, printers and so on. 4. Kitchen such as water boiler, toaster, mixer and so on. 5. Other devices such as iron, hair dryer, etc.
Table II. shows the distribution of household devices for single fulltime job family type [1].
𝑉𝐿 = 𝑉𝑆 −
𝑅𝑃+𝑋𝑄 𝑉𝐿
−𝑗
𝑋𝑃−𝑅𝑄
(2)
𝑉𝐿
TABLE II .Percentage of household devices in family category 1. Entertainment Beamer Recorder DVD Player Play Station WiFi
Communication 3% 41% 55% 5% 70%
Answering Cordless Ph.
Office 9% 23%
Kitchen
PC Laptop Fax Scanner Printer Hardware
PV systems has the analogy of a battery of very low voltage (around 0.6V), continuously recharged at a rate proportional to the incident illumination. Cells are connected in series-parallel, which allows the design of solar panels with high currents and voltages [2]. Solar PV systems are expected to become one of the most ideal renewable energy resources which can be implemented in large scale. [7]. Lately, literature surveys are done on how PV systems affect Power Quality. One of the most common PQ problems is Voltage Unbalance, raised due to rooftop PV systems in low voltage residential distribution system. A single phase rooftop PV model developed by MATLAB is used in the study for analyzing voltage unbalance under different installation ratings and location [3]. In this paper, rooftop PV systems are installed a single phase systems. When the rooftop PV is operated in a low voltage distribution system, a voltage profile will depend on the power output generated by the rooftop PV and the line impedance.Fig.1 is used to evaluate the effect
Fig.1. Rooftop PV connected to the grid.
of rooftop PV on voltage profiles [3]. In Fig.1, VSis the source bus voltage, VLis the load bus voltage and Z = R + jX is line impedance [3]. The load bus voltage can be derived by (1). 𝑅 + 𝑗𝑋 ×
𝑃−𝑗𝑄
(1)
𝑉𝐿
For the distribution overhead line, value of R is close 𝑋𝑃−𝑅𝑄 to X. If 𝑉𝐿 = 𝑉𝐿 ∠0°, the imaginary part of 𝑗 𝑉 is negligible and (1) becomes:
𝑅𝑃+𝑋𝑄 𝑉𝐿
83% 71% 64% 39% 67%
Vacuum Iron Sew Moisture Hairdryer Toothbrush
100% 85% 38% 22% 78% 44%
(3)
Where 𝑃 = 𝑃𝐿 − 𝑃𝑃𝑉 .When𝑃𝑃𝑉 is higher than PL and Q is very small, voltage rise is calculated as shown in (4).
PHOTOVOLTAIC ROOFTOP SYSTEMS
𝑉𝐿 = 𝑉𝑆 −
Coffee M/C Water Boiler Toaster Grill Mixer
𝑉𝐿 = 𝑉𝑆 −
The survey in Table II was conducted in 2010. 3
43% 41% 2% 17% 50% 10%
Others
𝐿
𝑉𝐿 = 𝑉𝑆 +
𝑅𝑃 𝑃𝑉 𝑉𝐿
(4)
Doing all the analysis, the final equations of current injected into the bus is calculated by (5) [3]. 𝑉𝑑 𝑉𝑞 𝐼𝑑 𝑃𝑃𝑉 1 (5) 𝐼𝑞 = 𝑉𝑑2 + 𝑉𝑞2 × 𝑉𝑞 −𝑉𝑑 𝑄𝑃𝑉 Where 𝐼𝑑 and 𝐼𝑞 are injection currents in d-axis and q-axis of the PV installed bus. 𝑉𝑑 and 𝑉𝑞 are the voltages in d-axis and q-axis of the PV installed bus. 𝑃𝑃𝑉 and𝑄𝑃𝑉 are active and reactive powers generated by rooftop PV. After calculating the powers, percentage of Voltage unbalance is calculated. Voltage unbalance is the maximum phase voltage deviation from the average phase voltage divided by the average phase voltage [8] as shown in (6). %𝑃𝑉𝑈𝑅1 𝑀𝑎𝑥 𝑉𝑎𝑚 − 𝑉𝑎𝑣𝑔 , 𝑉𝑏𝑛 − 𝑉𝑎𝑣𝑔 , 𝑉𝑐𝑛 − 𝑉𝑎𝑣𝑔 = × 100 𝑉𝑎𝑣𝑔 (6) 𝑉𝑎𝑛 + 𝑉 𝑏𝑛 +𝑉𝑐𝑛 Where, 𝑉𝑎𝑣𝑔 = 3 The flowchart shown in Fig.2 is used to calculate percentage of Voltage Unbalance [Reference]. Firstly, for analysis low voltage distribution system in a rural area is selected. The installation locations and rooftop PV’s ratings are randomly generated. Then the percentage of PV penetration is checked. If the value is less than 100%, the next step is carried out otherwise, installation locations and rooftop PV’s ratings are generated again. MATLAB is used to simulate the time domain simulations.. If the voltage rise is in the boundary of normal voltage, by using (6), the percentage of voltage unbalance is calculated. This process is repeated 1,000 times for various values.
4
FACTORS AFFECTING FLICKER IN WIND TURBINES
From a consumer perspective, voltage flicker is a serious power quality problem especially in wind turbines because their output power depends on wind speed[9]. In wind farms, flicker is caused by natural wind speed fluctuations such as gust, aerodynamic events such as wind shear (WS), and tower shadow (TS), turbulence
4.2 Tower Shadow Tower prevents wind to pass freely. Effect of this event is more obvious in downward than upwind turbines. For downward turbines, the output power fluctuations can reach average 10% peak to peak and maximum wind speed can change about 20%[4]. This event is modelled by mathematical formula [13,14] given in 𝑣𝑒𝑞 𝑡𝑠 =
(7)
3𝑣ℎ 𝑅2
2 𝑎 3 𝑙𝑛 𝑏=1 [ sin 𝜃 𝑏
𝑅 sin 𝜃𝑏 2 𝑥
2𝑅 2 𝑎 2
+ 1 − (𝑅 𝑠𝑖𝑛 𝜃
𝑏 +𝑥
]
2)
4.3 Blade Break Down If one of the blades breaks, the output power will fluctuate. The produced power is lower than usual and if the blade breaks completely the tower will disjoint after a short time[4]. 4.4
Gearbox Tooth Crashing When a tooth of gearbox crashes, the torque from driving shaft (turbine side) does not transfer to shaft (generator side) at a specific time. Since this fault repeats periodically, the fluctuation appears in output power[4]. Suppose one of N gears is broken, the transferred torque will be equal to 𝑁−1 𝑇0 , 0 ≤ 𝑡 ≤ 𝜏 𝑁 𝑇= 𝑁−1 0, 𝜏≤𝑡≤𝜏 𝑁 (8) 2𝜋 Where 𝜏 = 𝜔 and 𝜔 is the shaft speed. 4.5
Fig.2. Flowchart to calculate the percentage of voltage unbalance
intensity, malfunctions in mechanical part such as blade pitching error, tower’s oscillation, yaw control error, blades crash, gearbox tooth crash and switching phenomenon in inverters and generators[10-15]. 4.1 Wind Shear Friction between wind and earth effects on the vertical wind profile. Therefore, the mean wind speed increases with height. This event is known as Wind Shear. When blades reach to maximum height, they experience maximum wind speed and vice versa[4].
Blade Pitching In pitch controlled wind turbines, an electronic controller in a wind turbine governor senses the output power several times. If the output power level jumps to an unsafe level, a signal is generated which turns the blades out of the wind. When the power level is lower, they are turned back to catch the wind energy at the optimal pitch angle. Therefore, if the speed changes, the output power will be constant. Theoretical power that can be obtained from blades is expressed as below [16] 1 𝑃𝑡ℎ = 2 𝜌𝑆𝑉 3 (9) 𝑃𝑟𝑒𝑎𝑙 = 𝐶𝑝 × 𝑃𝑡ℎ (10) Where 𝐶𝑝 is the rotor efficiency. These are the main factors which affects flickers in wind turbines. 5
SIMULATION RESULTS
Low Voltage residential distribution system of a rural area is used in the analysis. This system has mix of residential and small commercial customers. This study system has three-phase line and single neutral line to supply electricity to customers. Fig.3 shows the voltage profiles of Phase A, Phase B, and Phase C when the rooftop PVs are installed [3]. Case 1 : Power demand of the load is at light load condition and % PV penetration is varied from 0 to 100%
Fig.3. Voltage Profiles of each phase when the rooftop PVs are installed.
% Voltage Unbalance
Fig.5. Percentage of voltage unbalance in Case2.
% PV Penetration Fig.4. Percentage of voltage unbalance in case 1.
of the feeder capacity. The rooftop PV’s installation is at only on phase A along the load of customers. Fig.4 shows the effect of the output power from the rooftop PV [3]. Case 2 : Power demand of the load is at light load condition and % PV penetration is varied from 0 to 100% of the feeder capacity. The rooftop PV’s installation is at phase A and phase B only along the load of customers. Fig.5 shows the effect of the output power from the rooftop PV [3]. Case 3 : Power demand of the load is at light load condition and % PV penetration is varied from 0 to 100% of the feeder capacity. The rooftop PV’s installation is at phase A, phase B and phase C only along the load of customers. Fig.6 shows the effect of the output power from the rooftop PV.
Fig.6. Percentage of Voltage Unbalance in case3.
6
CONCLUSION
In this paper, a probabilistic model is developed for building the load profile of a settlement including households has been introduced. Afterwards, the simulations of the rooftop PV’s installations in low voltage residential distribution system of a rural area are carried out. Simulation results show that percentage of voltage unbalance is increased rapidly with increased power output from the rooftop PV. Case 1 and Case 2 has similar results and the percentage of voltage unbalance is increased to exceed 2%. However, in Case 3, it is found that the percentage of voltage unbalance never exceed the standard limit and lastly we focussed on the factors which affect flicker in wind turbines. Aerodynamic faults such as WS and TS and mechanical failure such as gearbox tooth crashing, pitch control error and blade crashing are factors which lead to flicker in wind turbines.
7 [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
[11] [12] [13] [14] [15] [16]
REFERENCES S.Ghaemi and G.Brauner,”Stochastic model for household load profile,” in Symposium Energieinnovation, Austria,2010,pp.10-12. Raji& M.T Kahn. (2012, May). Analysis of distributed energy resources for domestic electricity users. Journal of energy in Southern Africa. [online]. 23(2), pp. 50-55. C.Pansakul and K.Hongesombut,”Analysis of Voltage Unbalance due to rooftop PV in low Voltage Residential Distribution System,”,in iEECON,Chonburi,2014,pp. 1-4. M.Fooladi and A.A.Foroud. (2015,July). Recognition and assessment of different factors which affect flicker in wind turbines. IET Journals.10(2),pp. 250-259. K.Barry.(2000,Sep.26).PowerQualityPrimer(1st)[Online]. Available:https://www.fer.unizg.hr/_download/repository/Power_ Quality_Primer_-_Barry_W._Kennedy.pdf Handbook of Power Quality,1st,.John Wiley & Sons Ltd.,New York ,2008,pp.1-642. A. Tripathi, J. Snadhu, K. Arunachalam and S.Biswas,” A step forward in précising building roof top areas for solar PV installations,” in ICSTM, 2015, pp. 532-537. IEEE Recommended Practice for Monitoring Electric Power Quality, IEEE Standard 1159-1995. C.D. Roger,F.Mark and H.McGranaghan (2012). Electrical Power Systems Quality.(3rded.) A.TascikaraOglu, M.Uzunoglu and B.Vural (2011).Power Quality Assessment of wind turbines and comparison with conventional legal regulations:a case study in Turkey. Appl. Energy.88(5).pp.1864-1872. Y.Zhang and W.Chen(2014).Flicker mitigation strategy for a doubly fed induction generator by torque control.IET Renew. Power Generation.8(2).pp.91-99. D.S Dolan, P.W Lehn,”Simulation model of wind turbine 3p torque oscillations due to wind shear and tower shadow,”,inPSCE’06,2006,pp. 2050-2057. S.Das, N.Karnik and S.Santoso ,”Time domain modelling of tower shadow and wind shear in wind turbines,”,inInt. Scholar Research Note,2011,pp. 1-11. M.M Gilbert.(2013, Aug.).Renewable and efficient electric power systems.(2nd). D.S Chavan, P.B Karindakar and A.K Pande ,”Assessment of flicker owing to turbulence in a wind turbine placed on a hill using wind tunnel,”,inICCPT,Nagercoil, India,2014,pp. 560-566. M.Singh and S.Santoso,”Dynamic models for wind turbines and wind power plants,”,in National Renew. Energy Lab,January 2008May2011.
AUTHORS Principal Author: Gunjan Gupta received the B.Tech degree in Electronics & Telecommunication from U.P. Technical University, Lucknow,Uttar Pradesh, India in 2006 and M.Tech degree in VLSI Design U.P.Technical University, Lucknow,Uttar Pradesh, India. She had been with RKGIT, Ghaziabad affiliated to U.P.Technical University, Lucknow, Uttar Pradesh, India as an Assistant Professor for 9 years and is currently pursuing her Ph.D from Cape Peninsula Univeristy, Cape Town, South Africa. Co-author: Dr.Wilfred Fritz is a senior lecturer at the Department of Electrical Engineering at CPUT, hold SA BEng ,two masters and D Tech degrees from the universities of Stellenbosch, Paris and CPUT respectively. He is a registered Certified Measurement and Verification Professional (CMVP) and also a professional engineer (PrEng) affiliated to the Association of Energy Engineers (AEE) and ECSA respectively. He has a passion for renewable energy and energy consumption reduction technologies.
