Design and Development of Grid Assisted Adaptive Solar (PV)

Document Sample
Design and Development of Grid Assisted Adaptive Solar (PV) Powered By Docstoc
					                                ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                    Providing Opportunities for Technological Advancement


              Design and Development of Grid Assisted Adaptive
   Solar (PV) Power Converter for Water Pumping System in Indian villages
                             Shiva Nand Singh *, Arun Kumar Singh **
           Department of Electronics Engineering*, Department of Electrical Engineering **
                        National Institute of Technology, Jamshedpur, India
                          * snsnitjsr@gmail.com **aksnitjsr@gmail.com

Abstract
The paper presents the development of a grid assisted solar (PV) power converter for a water
pumping system used in rural homes of Indian villages. The power supply system is composed
of a solar (PV) array, PWM inverter incorporating PWM control strategy, energy storage
batteries, self priming pump and water storage tank(s). The model of the system was designed
for optimal operation, and a prototype of the system was developed to produce 500VA power
with low Total Harmonic Distortion (THD) to drive a ½ HP pump-motor and provide power for
lighting. A life cycle cost evaluation for service period of the system was done and compared
with a conventional Diesel Generator set. A cost effective system with around 40% grid power
saving was achieved.

Keywords: Pulse Width Modulation, Total Harmonic Distortion, Photovoltaic, Solar module

1. Introduction
Water is a basic need of all living beings. About 30% of the world population lack access to
water for drinking, live-stock and irrigation. Traditional technology used to access water from
available sources like bore well or open well employ a water pump to lift water and store it in
tank(s). These pumps are either powered by conventional grid supply or alternative Diesel
Generating (DG) set. More recently, higher cost of fuel consumed by DG set and non availability
of adequate grid supply have forced scientists and engineers to think of alternative or
supplementary sustainable renewable solar energy sources [1] to produce electrical power. The
fuel (i.e., sunlight of solar renewable energy source) is abundant in nature, is free and produces
green electricity. Solar photovoltaic conversion technology [2, 3, 4] is also simple and straight
forward. It is also gaining more popularity due to its other features like long life, lower
maintenance and reliable operation. Although the solar modules (cells) are expensive, efforts
are being made to utilize such modules not only for power conversion, but also for building the
exterior wall or covering the roof of water pump houses.

Demand for electricity is increasing day by day due to growing population where grid supply
extension has almost come to a standstill due to limited resources like fossil fuel cost increase
and various techno-economic reasons. This has motivated researchers to develop grid assisted
solar power converters [5, 6, 7] to generate power which can meet the increasing energy
demand of houses located in rural sectors of the country connected to weak grid supply sources.

2. Methodology (Modeling and Design of Water Pumping System)
Modeling and design of water pumping system include selection of modules (Refer to Figure 1).
• PV cell module
• Battery storage and back-up
• PWM inverter
• Intelligent power controller

             Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                    2009-4-01-REEF
                              ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                  Providing Opportunities for Technological Advancement


•   Pump for water extraction and storage in overhead tank etc.




                 Figure 1: Model of solar (PV) powered water pumping system

Data on electric power demand load profile were assessed. The computational analysis for
optimal design of the system components was carried out to develop a prototype inverter model
and to test its dynamic performance. The proposed system is able to achieve an energy saving
up to a maximum value of 40-45 % of power drawn from utility supply in these rural houses.
The present study covers the following:
• Study of user demand of pump and lighting load profile in a rural house of Indian villages.
• Optimal design of solar converter system including PV array, batteries, PWM inverter and
    pump-motor.
• Prototype development of utility interfaced adaptive intelligent power controller unit.
• Economic impact analysis of the prototype system and its life cycle cost comparison for
    service period with Diesel Generator set.
• Social impact of using the system on the rural development in Indian villages.

The solar energy is harnessed through photovoltaic cell and converted into utility grade AC
power using PWM inverter. The converter works in bi-directional mode and performs both under
charging and inverter mode of operation. The push pull configured centre tapped converter unit
is switched ON and OFF alternatively in push pull mode by transistorized switching power
devices T1 and T2 to produce ac power from the DC. (Figure 2)

The system adapts to meet the varying load (Lightning and Pump) under three modes of its
operation

Mode 1:    High insolation with more than 50% state of charge (SOC) of the battery: The PV
           shares power with the grid and charges the battery bank as well as feeds load(s)
           through PWM inverter. Switch S1 is made OFF

Mode 2:    No insolation (SOC more than 50%): Battery shares power with the grid and feeds
           power to load(s) through PWM inverter. Switch S1 is made ON
             Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                    2009-4-01-REEF
                               ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                   Providing Opportunities for Technological Advancement




Mode 3:   Low battery condition (SOC less than 50%): Battery is disconnected from load. PV
          and/or grid charges the battery till it becomes fully charged and simultaneously feed
          power to load through inverter. Switch S1 is made OFF




            Figure 2: Circuit model of utility interface solar power converter system


The switching of these power devices are controlled through PWM base drive pulses at base of
Transistors T1 and T2 generated by Base drive module (Figure 1) and thus minimizes the power
loss. The intelligent controllers monitor and control the system parameters and perform the
various power management control tasks such as:

•   PV power management
•   Battery management
•   Load power management

3. Optimal Design of System Component
3.1 PV Sizing
The empirical formula based on energy balance equation [8] has been used to compute the
optimal size of PV module for critical limit of load as stated below:

       PV Cell Rating (PPV) = (PTL * S.F) / sun hour [watt]                                (1)

       Where ,

       Sun hour = 6.2 for adopted area
            Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                      2009-4-01-REEF
                               ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                   Providing Opportunities for Technological Advancement


        Safety Factor (S.F) = 1.5 for cloudy weather
        PTL is total load energy in watt- hours (i.e. total load power over a period of 24 hours
        assuming hourly load power (PL) as constant.)

                                   24h
                       PTL (Wh)   = Σ (PL) Watt-Hours                                            (2)
                                   0h

The optimal number of PV module = PPV / Standard PV module rating                                (3)




3.2 Battery Sizing
The battery stores the energy to a maximum value as per average load power requirement.

                              Battery capacity (Ah) = PTL / (12V * SOC)                    (4)

Where
                              SOC (State of Charge) of Battery = 50%

3.3 Pump Motor
The pump is driven by ac motor whose optimal value can be computed by the following
expressions.
                                                     HP
                              Motor Power =                                                (5)
                                                     η
Where
                              HP = Hydraulic power of pump [W]
                              η = Efficiency of pump

The hydraulic power HP can be computed by the following expression
                            HP = Q ρ g H                                                         (6)
Where
                            Q = discharge rate m3/s
                            ρ = density of water 1000kg/m3
                            g = acceleration due to gravity 9.81 m/s2
                            H = Dynamic Head (m)

3.4 Inverter Module

The inverter produces AC power output with DC power input .The efficiency of Inverter depends
on power losses occurred due to inverter components like transformer, switching losses in
power devices (i.e. Transistors) and harmonic content in ac output power. The PWM pulses
generated through microprocessor/computer software program are approximated to a sine wave
(Figure 3) resulting in efficient utilization of transformer and low power loss in power switching
devices of inverter.


              Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                     2009-4-01-REEF
                              ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                 Providing Opportunities for Technological Advancement



4. Inverter DPWM drive algorithm
The system controller produces PWM pulses using Direct PWM modulation (DPWM) control
strategy (Figure 3). The pulse width (Pi) of DPWM approximating to an equivalent sine wave is
expressed as stated below:

                                                                                             (7)
Where
                   Pi = PWM pulse width of ith pulse
                   N = number of PWM Pulses within half cycle approximating sine wave
                   i = 1, 2, ..., N




    Figure 3: Simulated Direct modulated sinusoidal PWM pulses for N = 3 in one half cycle
         (10ms=180 Deg), Frequency=50 Hz (Scale: X=degree, Y=pulse voltage x 5V)

5. Hardware implementation
The software algorithm generating the number of pulses N of PWM Pulses has been
implemented through 16 bit Microprocessor (8086). The pulse width and Notch width timings
are computed from the switching angles of PWM pulses and are loaded in the timer (peripheral
device) unit of Microprocessor and brought out through ports interfaced with Microprocessor.
The program flow-chart is as depicted in Figure 4.

                                           START



                                  Initialize 8253 and 8255
                                   Microprocessor (8086)



                                   Compute Pi and Ti from
                                    the Switching Angle



                               Loading of Computed Value of Pi
                                and Ti in 8253 and brought out
                                         through 8255

                                                                   YES

                                             Is i = N
             Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                    2009-4-01-REEF
                                  ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                      Providing Opportunities for Technological Advancement




                     Figure 4: Flow chart of PWM pulse generator program
6. Testing Specification
A prototype solar converter unit of the system was designed and developed as indicated in
Table 1:



        Table 1: Technical Specification of Prototype Sample of Solar Converter system

            PV Cell                    12 V, 75 Wp @ Standard Test Condition (STC)
      Battery (Lead Acid)                                12 V, 150 Ah
        PWM Inverter                                 500 VA, 220 V, 50 Hz
                                     Pump (self priming 1/2 HP centrifugal surface pump for
              Load
                                  dynamic head of 10 meter, Compact Fluorescent Lamp (CFL)
          Grid Power                                  220V ±10% , 50 Hz

7. Adaptive control of Power Flow and Response of System
7.1 Power Flow Control Algorithm
The adaptive model of power flow control system is governed by equation 8. The Power flow
control algorithm is governed by conditions set in Table 2 [8].

                                 PL = PPV ± PBAT + PGRID                                            (8)
Where
                                 PL = Load Power [W]
                                 PBAT = Battery power [w]
                                 PGRID = Grid rectified DC Power [W}
                                 PPV = PV power [W]

                             Table 2: Digital adaptive power controller
                                   Energy Source(s)
                                                                                            Load
  Power                                        Battery
                Grid        PV                                                    (-ve sign power drawn)
  Status of                         (+ve charging, -ve discharging)
  Energy
  Source          0         0                          +                                      -
  & Load          0         1                         +/-                                     -
                  1         0                         +/-                                     -
                  1         1                          +                                      -

7.2 Response of the System
The adaptive digital logic controller (ADLC) integrates the power sources and gives a consistent
output power as per load requirement. The system response time under switching/changing
load, grid variation, varying solar insolation have been modeled for less than 3 ms with an
overshoot and undershoot limited to 1.5 % and settling time within two cycles of power
frequency.


              Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                     2009-4-01-REEF
                              ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                   Providing Opportunities for Technological Advancement




                             Figure 5: Response of solar converter

8. Experimental Investigation and Observations
8.1 Data Acquisition of Solar Radiation
The solar power radiation data are measured with the help of pyrometer during sun hour period
falling on PV module of 75 Wp and is recorded as given in Table 3:

Table 3: Solar Radiation Data (power out put) of solar cell acquired during Sun Hour Period on a
typical day in the month of July 2007(Total Power 523 Watt-Hours)

                       Time (AM)      6       7      8      9       10     11
                       Power [W]      0       5      50     60      70     73
                       Time (PM)      12      1      2      3       4      5
                       Power [W]      75      65     60     55      10     0

8.2 Load Power Profile
The demand base peak load power of a farm house (lighting and pump motor) profile
encountered by the system has been recorded and plotted in Figure 6 .The users fix up priority
to switch ON lighting loads connected at various locations of farm house and also the operating
time and its duration for pump load.




 Figure 6: Peak load power demand profile of a rural house for a typical day (24Hour) recorded
                                 in the month of July 2007
             Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                      2009-4-01-REEF
                              ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                  Providing Opportunities for Technological Advancement


8.3 Computational Design Analysis
The system component parameters i.e. PV sizing, battery sizing were designed for a proto type
sample using window based software MS Excel Work-Sheet. A prototype sample is shown in
Figure 7.

Load Profile =   1200 watt-hours over 24 hours
PV size      =   75 Wp, 12 V
Power shared =   45% (approx) of total power produced by PV and balance shared with grid
Battery Size =   150 Ah, 12 V lead acid for load power and 7.5 Ah for control power supply
Pump Motor =     1/2 HP, self priming, Dynamic head 10 -15 meter, 3000 liter Tank
Inverter     =   500 VA, 220 V PWM AC ,50Hz




         Figure 7: Prototype sample of utility interface adaptive solar power converter

8.4 Inverter Output Waveform
The image of PWM base drive pulses and Inverter output waveform on oscilloscope are shown
in Figure 8 (a) & (b).




   Figure 8: (Left) PWM control drive pulse, (Right) Inverter Output waveform (220V, 50Hz)




             Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                    2009-4-01-REEF
                                 ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                      Providing Opportunities for Technological Advancement




8.5 Harmonic Analysis
The harmonic content of the PWM pulses for different value of N( = 3…11) number of pulses per
half cycle is computed using MATLAB Program as a total harmonic distortion (THD) up to nth
i.e 15th harmonic (considering as an objectionable limit in motor driven pump applications) as
shown in Table ( 4 ). From the table, it is evident that with the increase in number of PWM pulse
(i.e N ) the waveform approximated towards near to sine wave and thus resulting in decreasing
THD value at nth value of harmonic.

                               Table (4): % Harmonic analysis (THD)

             N
                      n th    N=3       N=5           N=7             N=9                N=11
                 3           0.1931    0.0594        0.0387          0.0250             0.0384
                 5           0.4462    0.0628        0.0419          0.0460             0.0435
                 7           0.4571    0.1668        0.0441          0.0493             0.0463
                 9           0.5731    0.3452        0.0500          0.0494             0.0488
                 11          0.5750    0.3607        0.1890          0.0520             0.0501
                 13          0.5751    0.4223        0.3162          0.0594             0.0528
                 15          0.6151    0.4680        0.3359          0.2111             0.0556


8.6 Efficiency of converter
The Efficiency of the solar converter has been computed using the observed value of input and
output power of the inverter. The plot of Efficiency v/s normalized load is shown in Figure 9,
which shows almost constant efficiency with an optimum high value at 30% and low value at
90% of rated load.




                             Figure 9: Efficiency Graph of solar converter

8.7 Battery Charging
The charging of the battery by the solar power as well as from the utility supply during sun hour
period and night hours respectively under two modes of its operations [9] :
• Trickle mode(low current): Under low Insolation condition the terminal voltage is set as 2.2V
    - 2.35 V per cell


             Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                    2009-4-01-REEF
                                 ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                     Providing Opportunities for Technological Advancement


•  Bulk charging mode (high current): under high insolation condition the terminal voltage of
   battery is set as 2.5 - 2.7 V per cell.
The charging current drawn by the battery at standard C-10 rate of charging and the status of
terminal voltage is recorded as shown in Figure 10.




                                   Figure10: Charging of Battery

9. Economic Impact Analysis
Although the initial cost of solar (PV) converter system as compared with DG set is high but its
Life Cycle Cost (LCC) is significantly lower. The cost estimation of prototype unit of specified
range of 300 - 500VA has been done for a service life period of 3 years over life cycle of 20 -30
years and comparison with conventional DG set has been done (Table 5) to validate the same.
Table 5 shows that saving in cost for a service life period of 3 years is approximately 50%.

    Table (5): Cost analysis and comparison for service life period of 3 years of a prototype unit
                                of solar power converter and DG Set

Solar (PV) converter system module                 (cost in Rs.)          DG set            (cost in   Rs.)
              PV array 75 Wp, 12 V                         20000            Generator 1/2 HP           18000
     Battery 150 Ah / 12V (3 yrs service life)             10000         Fuel (K.oil + Petrol) Rs
                                                                         50 per day (max) for 3        55000
          Inverter (300- 500 VA, 220 V)                     5000
                                                                        year and operational cost
                  Maintenance                               1000              Maintenance              5000
Grid supply cost Rs 5 per day (fixed) for 3 years           4000
                      Total                                40,000                    Total             78,000

10. Socio- Economic Impact of system on Rural Development
The study revealed that dependency on grid power supply required for meeting the energy
demand - a basic need for growth and development of rural India, can be reduced to minimum
or almost negligible with the use of more solar panel. The cost of the system can also be made
comparable if PV modules are used as building material in constructing wall or roof of rural farm
houses. The supplementary or alternative source of solar energy can bring green environment
by providing electricity and water for drinking and irrigation in every rural house resulting into
increasing the status of a few parameters of socio- economic developmental activities such as :
• Education (Literacy )
• Heath, Hygiene and sanitation
• Income generation
• Quality of life
               Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                      2009-4-01-REEF
                               ARAB RESEARCH INSTITUTE FOR SCIENCE & ENGINEERING
                                   Providing Opportunities for Technological Advancement




11. Conclusion
In this paper, investigations have been carried out on the development of solar power converter
as a supplementary source to conventional grid for agriculture and household applications for
rural masses in Indian villages. The following features have been included in the system

•   The technology used in the proposed scheme is simple, cost effective and having fast
    response in terms of control stability under varying solar insolation and/or agriculture load
    conditions.
•   The system can be easily scaled to higher ratings. It is easily to maintain and conforms to
    sinusoidal quality with low switching loss and with low harmonic content.
•   The load power adaptability feature due to integration of Input power sources (i.e. PV,
    Battery & grid ) and PWM control strategy used in inverter offer consistency in power supply
    irrespective of varying level of load demand(s) and /or solar insolation.
•   The system offer almost constant efficiency with high value and produces grid quality power
    with low distortion (THD )value
•   This technology can play a very vital role on rural development in terms of agricultural
    productivity, economic status and standard of living.

Acknowledgment
The authors wish to thank the support of the Director, National Institute of Technology,
Jamshedpur for completing the research work.

References
[1] htpp:/ /www.pvpower .com
[2] Mahmud, (1998) “Solar PV submersible water pumping for small scale use in Bangladesh”,
     M.Sc Engg thesis, Mechanical Engineering Department, BUET, Supervisor Prof. Dr. M..
     Rashid.
 [3] http://solstice.org Conference (PVSEC 2000) Glasgow (Scotland) May 2000, page 2422-
     2425.
 [4] G.J. Vander Metwe, ‘150W Inverter – an optimal design for use in solar home system’,
     International Symposium on Industrial Electronics Proceedings of ISIE, 1998, Vol. 1, page
     57-62.
 [5] C. Alonso, “Design of Optimized Photovoltaic Conversion Chains”, 16th Photovoltaic Solar
     Energy source inverter” First WCPEC, Dec 5-9, 1994, Hawaii.
 [6] B. Lindgrin, ‘A 110W inverter for photovoltaic application’, Published in International
     Journal of Renewable Energy Engineering, April 2002
 [7] www.statpower.com Cole, N. (1995), “Renewable are ready - peoples creating the
     renewable energy solutions”, Union of concerned scientists.
 [8] S.N.Singh, A.K.Singh “Modeling and simulation of PV powered solar power converter for
     use in water pumping application ‘journal of ieema, Jan 2008 p114-122
 [9] Wang Jian, “Optimal control of Solar Energy combined with MPPT& Battery Charging”
     Proceeding of Sixth International journal conference on Electrical Machine & System 2003
     ICEMS2003 page 285-288.




             Ariser: Journal of ARISE, ISSN 1994-3253, www.arabrise.org
                                    2009-4-01-REEF