IMPROVING THE EFFICIENCY OF SOLAR PHOTOVOLTAIC POWER SYSTEM

As the local and national clamor for foreign energy independent United States continues to grow unabated; renewable energy has been receiving increased focus and it’s widely believed that it’s not only the answer to ever increasing demand for energy in this country, but also the environmentally friendly means of meeting such demand. During the spring of 2010, I was involved with a 5KW solar power system design project; the project involved designing and building solar panels and associated accessories like the solar array mounts and Solar Inverter system. One of the key issues we ran into during the initial stage of the project was how to select efficient solar cells for panel building at a reasonable cost. While we were able to purchase good solar cells within our allocated budget, the issue of design for efficiency was not fully understood , not just in the contest of solar cells performance , but also in the overall system efficiency of the whole solar power system, hence the door was opened for this thesis. My thesis explored and expanded beyond the scope of the aforementioned project to research different avenues for improving the efficiency of solar photovoltaic power system from the solar cell level to the solar array mounting, array tracking and DC-AC inversion system techniques.


LIST OF TABLES
Coal, Petroleum and Natural gas, coupled with the environmental degradation caused by the process of harnessing these energy sources, it has become an urgent necessity to invest in renewable energy resources that would power the future sufficiently without degrading the environment through green house gas emission. The energy potential of the sun is immense, but despite this unlimited solar energy resource, harvesting it is a challenge mainly because of the limited efficiency of the array cells. The best conversion efficiency of most commercially available solar cells is in the range 10-20s% [1], [8]. Although recent breakthrough in the technology of solar cells shows significant improvement but the fact that the maximum solar cell efficiency still falls in the less than 20s% range shows there are enormous room for improvement. The goal of this thesis is to identify these rooms and ways to improving them. One of such room is array mounting and tracking mechanism that moves or positions solar array to absorbing extended solar irradiance for maximum power output. Another such room is researching different types of solar cells from past to present and the future trend and identifies the sources of losses and how to mitigate them. Lastly, some critical components that are necessary for efficient operation of solar power inverter system are investigated.

REVIEW OF SEMICONDUCTOR DEVICE PHYSICS OF SOLAR CELLS
Solar Cell Device Physics:

Figure 1.1 Typical Crystalline Silicon (cSi) Solar Cell [1], [15]
Solar cell like the crystalline silicon based solar cell shown in Figure 1 above is a solid state semiconductor p-n junction device that converts sunlight into direct-current electricity through the principle of photo-voltaic effect. The first conventional photovoltaic cells were produced in the late 1950s, and were principally deployed to provide electrical power for orbital satellites. During this initial deployment, excessive cost of manufacturing and poor efficiency of solar modules were some of the major challenges that limit their competitiveness as a major source for meeting the increasing energy demand that has continued till now. However, recent improvements in design, manufacturing, performance, reduced cost and quality of solar cells and modules have not only opened up the doors for their deployments in applications like powering remote terrestrial applications, rural electrification projects, battery charging for navigational aids, water pumping, telecommunications equipment and critical military installations, but has also propelled solar power system as a competitive means to meeting the ever increasing power need for the world economy. While the focus of this thesis is improving the efficiency of a solar power system, it's important to take some cursory, but refreshing look at the semiconductor physics of a solar cell. The PN-Junction is the region or a boundary that is formed by doping or by epitaxial growth of a layer of crystal doped with one type of dopant on top of a layer of crystal.

P N P N
A dopant is a material (Impurity) that is intentionally introduced or mixed into an extremely pure (Intrinsic or un-doped -Crystal Silicon for example) semiconductor material for the purpose of changing or optimizing its electrical properties for specific application. A dopant could be n-type or p-type material. Thus, a P-type Doping is the introduction of impurity atoms with one less valence electron-(Like Boron) than silicon (acceptor impurities), resulting in excess positive charge carriers (holes).
Whereas, an N-type doping is the introduction of impurity atoms with one more valence electron (like Phosphorous) than silicon (donor impurities), resulting in excess negative charge carriers (electrons). electrons from the high concentration n-type side tends to flow towards the p-type side and similarly the holes from the high concentration p-type side tends to migrate towards the n-type side. These electron and holes migration creates charge imbalance by exposing ionized charges on both sides. The exposed charges would set up an electric field that opposes the natural diffusion tendency of the electron and holes at the junction; this is the behavior of the PN-Junction under equilibrium condition.
Under this condition, there exists within the junction a layer between the PN-Junction that becomes almost completely depleted of mobile charge carriers. This layer/region is called the space-charge region or depleted region and is schematically illustrated in     If a load is connected between the electrodes of the illuminated p-n junction, some fraction of the photo-generated current will flow through the external circuit. The potential difference between the n-type and p-type regions will be lowered by a  The PN-Junction IV characteristic curve of an ideal diode solar cell is described by the Shockley equation 3 below [1]; The Shockley equation is the fundamental device physics equation which describes the current-voltage behavior of an ideal p-n diode. I photon is the photo-generated current and is defined by equation 4 below, [1].
Where L e and L h as defined before are the minority carrier diffusion lengths for electron and holes respectively. G is the diode electron-hole pair generation rate, W is the width of the depletion layer and A is the total illuminated cross sectional area of the device. Based on this equation, it can be inferred that only carriers generated in the depletion region and in the regions up to the minority-carrier-diffusion length from the depletion region contributes to the photo-generated current.

Solar Cell Conversion Efficiency η:
The conversion efficiency of a typical solar cell is the ratio of the maximum output generated power to the input or incident power. Certain output parameters greatly influences how efficient a solar cell is and are defined as follows. can deliver a maximum of 46mA/cm 2 under an AM1.5 spectrum [1].

Open Circuit Voltage (V oc ):
The

Fill Factor FF:
The fill factor FF is the ratio of the maximum power (P MP) generated by the solar cell to the product of the voltage open circuit V oc and the short circuit current I SC From the above set of equations we define the solar cell conversion efficiency η as the ratio of the maximum generated power (P MP =V MP .I MP ) to the input or incident power Pin as given by equation 10 below.
Pin is the total power of sunlight illumination on the cell. Energy-conversion efficiency of commercially available solar cells typically lies between 10 and 25 % [8]. These three important parameters (V oc , I sc and FF) as described above are the most important factors that determine how efficient a solar cell is and are optimized for efficient solar cell design.

Improving the Conversion Efficiency of Solar Cell:
This section identifies the major sources of loss in the solar cell conversion efficiency process and the corresponding approaches to mitigating the losses thereby improving the efficiency.

Light Energy (Photons) Absorption:
Sunlight is a portion of the electromagnetic radiation (Infrared, Visible and Ultraviolet lights) that is emitted by the Sun. On Earth, sunlight is filtered through the Earth's atmosphere, and is visible as daylight when the Sun is above the horizon. The amount of radiant energy received from the Sun per unit area per unit time is called Solar Irradiance and it is a function of wavelength at a point outside the Earth's atmosphere.
Solar irradiance is greatest at wavelengths of between 300-800 nm. The AM 1.5 spectrum which correspond to Latitude 48.2˚ is the preferred standard spectrum for solar cell efficiency measurements. Where Lo (the zenith path length) is perpendicular to the Earth's surface at sea level and θ is the zenith angle in degrees.
The air mass number is dependent on the Sun's elevation path through the sky and therefore varies with time of day and with the passing seasons of the year, and also with the latitude of the observer. At the outer space i.e. beyond our terrestrial environment, the solar spectrum has an Air Mass coefficient of zero (AM0) as seen in The question then is how many photons can be absorbed per unit area of solar array?
Let us consider silicon cell as example to answer this question; if we have a Silicon solar cell with energy band gap of E= 1.1eV. Only photons with energy greater than 1.1eV and wavelength λ < λ bandgap , about 1.13um would be absorbed and the rest will be lost as heat [28]. Also, even when the incident light with the adequate energy level and wavelength strikes the surface of the cell material, some photons are reflected from the surface of the solar cells; All these leads to reduced efficiency.
One way to ensure maximum absorption is through the use of cell material with very low reflective coefficient or placing a thin film anti-reflective coating over cell surface. Another method that can be employed in reducing reflection is using textured surface, in which the direction of reflected light on the textured surface is downward so that reflected photons can be reabsorbed again by the cell, thereby improving the conversion efficiency [28].

Losses Due to parasitic resistance:
Another source of loss is through parasitic resistance. The equivalent circuit of practical solar cell is shown in the        Increasing the azimuth angle maximizes afternoon energy production. For a fixed PV array, the azimuth angle is the angle clockwise from true north that the PV array faces and for a single axis tracking system, the azimuth angle is the angle clockwise from true north of the axis of rotation. The azimuth angle is not applicable for dual axis solar tracking PV arrays. [5]

Table 1.2 Azimuth Angle by heading [25]
The sun's height above the horizon is its altitude and it changes based on time and season of the year. Based on the sun's altitude changes, the tilt angle of a solar module with respect to the sun must be carefully considered during module or array installations. The general practice for fixed array is that the tilt angle be equal to the latitude, for providence Rhode Island, the latitude is 41.73˚. For better absorption, it is recommended that the tilt angle be adjusted to Latitude + 15˚ during winter and Latitude -15˚ during summer.

Solar Tracking
A solar tracker is a device that move or adjust the positional angle of solar photovoltaic panel towards the sun. The sun's position varies both with season and time of day as the sun moves across the sky. Solar panels absorb energy better when orientated perpendicular to the sun, therefore the solar tracker mechanism will essentially increase the effectiveness of solar panels over a fixed solar array or panel.

Types of tracker
Trackers can be categorized by the complexity of operation and sophistication.
There are two major groups; Active and passive Trackers. Passive trackers are without motor. Active trackers are motorized and can be sub-categorized into single axis and dual axis trackers:

Single axis
Solar trackers can either have a horizontal or a vertical axis. The horizontal type is used in tropical regions where the sun gets very high at noon, but the days are short.
The vertical type is used in high latitudes where the sun does not get very high, but summer days can be very long. In concentrated solar power applications, single axis trackers are used with parabolic and linear Fresnel mirror designs.

Dual axis
The Dual axis solar trackers have both a horizontal and a vertical axis and thus they can track the sun's apparent motion virtually at any angle. A dual axis tracker maximizes the total power output of solar array by keeping the panels in direct sunlight for the maximum number of hours per day.

Tracker Components:
A typical solar tracking system consist of mechanical parts like the linear actuator with integrated dc or ac motor and gear and an electronic parts like motor drive and controller, sun sensor and power supply. Based on complexity of design and accuracy of tracking, there might be more additional components than those mentioned above. For example, The AZ125 Wattsun dual tracker shown in Figure   1.19 below can move up to 270˚ in the horizontal direction and up to 75˚ vertical [23].

Fixed array Vs Tracked array
A typical solar array with polycrystalline solar cells with 14-18% conversion efficiency would need area of about 7-8m 2 to produce a 1KW peak power. Some applications does not allow enough space for larger array area but still need to produce sufficient power to meet a desired energy need. Therefore the question of whether to track an array or not is based on specific situation and need. For a fixed Array, the general norm is to install at angle of the Altitude. In the case of Rhode Island region, the Altitude is approximately about 41degree. However, in one of my evaluation of the tilt angle adjustment for a fixed array, I found out that mounting an array at an optimum angle of mount for a stationary array would generate more power than just using the Altitude angle all year round. Figure 1.20 below shows a 30W array that was mounted at the altitude angle as well as at angle slightly higher than the altitude angle under similar environmental conditions during the beginning of winter 2012.
The results as shown in the plot of Figure 1.21 shows that mounting at angle slightly higher than the altitude (blue) has about 5% peak power improvement over the altitude tilt angle (green).

PV-WATTS Modeling:
To further demonstrate the advantage of tracked array over fixed array, the 5KW power system design was also modeled using the PV-Watts simulation tool, which is an Internet-accessible simulation tool for providing quick approximation of the electrical energy produced by a grid-connected crystalline silicon photovoltaic (PV) system for up to about 239 locations in the US. Users would typically select a location from a station map and set the PV system parameters and the PVWATTS performs an hour-by-hour simulation of monthly and annual alternating current (AC) energy production in kilowatts and energy value in dollars. Some of the system parameters that may be specified include AC rating/size, local electric costs, PV array type (fixed or tracking), PV array tilt angle, and PV array azimuth angle [29]. The tool is available at the National Renewable Energy Laboratory (NREL) website [10]. The results as shown below; proved that the tracked array generates more power (22.8% more) than fixed array.  Apart from installing an array for optimum tilt angle, array sizing is also very important when deploying a solar power system. Photovoltaic power system are sized based on ; load estimate , inverter system power estimate, wiring sizes, battery and the charge control size estimate (for hybrid or standalone system) and available area. In the 5KW power project for example, (one of the panels we built as shown in Figure   1.24 below) space was not an issue, therefore tracking was not considered. The load demand analysis for the project is as follows and also detailed in table 1.2 below Load Analysis for the 5KW power system: From the table, the size of the modules was based on; The total daily Amp-Hours/day = 1737 Average Sun-hours/day = 5 based on location in Newport RI; Minimum Power required to meet demand = = 4.168KW Because of system losses, it is a general practice to add 20% to minimum power when sizing array; the array size was therefore based on 4.168KW + 0.2*4.168KW = 5KW.
A 12V/294W module would require 17 parallel panels to make 5KW array as shown in the schematic Figure 1.23 below.  One key requirement of all grid interactive inverters is that the inverter must be able to recognize and stop operating if the grid is not present either due to fault or in case of utility scheduled service based on IEEE-1547 standard requirements. A bi-directional interface and feedback compatibility is made between the PV system AC output circuits and the electric utility network through the AC disconnects switch and the contactor. This interface also allows the AC power produced by the PV system to either supply on-site electrical loads or to back-feed the grid when the PV power system output is greater than the on-site load demand. At night and or during overcast situation when the electrical loads are greater than the power output from the PVI, the balance of power required by the load is received from the grid.

Smart Grid Photovoltaic Inverter System (SGPVI):
There are tremendous research going on to refurbish the grid of the future for smartness and efficiency. The 'Smart Grid' concept is still evolving and so also are the standards that defines and qualifies what a smart grid is. Because these standards are yet to mature but still in the developmental stage, it is difficult to ascribe a specific definition to smart grid. However, Smart Grid in the concept of the integration of the electrical and information infrastructures into the utility system can be defined as the incorporation of automation and information technologies with the existing electrical network in order to provide overall improvement to the utility's power reliability and performance. The smart grid would function to deliver increased energy efficiencies and provide significant reduction in carbon emissions and other environmental hazards. It must also provide consumers with the flexibility to manage their energy usage to save money on energy while supporting renewable energy integration.
Additional advantage of the smart grid is that it must be intelligent to diagnose power outage and troubleshoot to provide automatic restoration. It must also be able to provide security of the power network system by detecting and reporting network system intruders [30]. Smart Grid total solutions is being driven to provide Assets and Demand Optimization, Smart metering and communications, Distribution and Transmission optimization, and Engineering design optimization [30]. The Smart grid inverter system must therefore be designed to meet these new grid requirements as specified in the IEEE-1547 standard (Physical and electrical interconnections between utility and distributed generation (DG)) and other standards like ANSI C12.19/MC1219, Advanced metering infrastructure (AMI) etc.

Grid-Integrated Photovoltaic Inverter System:
The process of converting the dc power from the solar array to ac power which can then be used to power industrial or household ac loads or interfaced into the grid is known as inversion. Other than dc-ac inversion, the other major functions or features that are common to most grid-tied solar inverter systems are;  Maximum power point Tracking,  Grid integration and disconnection. to convert the ac voltage to the correct grid ac voltage both in magnitude and phase angle. The transformer also serves to provide isolation between the grid and the PV.
The DC input voltage is typically about 300 -600Vdc in the US and can go up to about 1KV in Europe.

Figure 1.26 2-level 3-phase Inverter Topology [18]
Multi-level Inverter Topology: Another type of inverter topology is the multi-level inverter.   Therefore, Q2 and Q3 would have greater conduction loss than Q1 and Q4 but far less switching loss. This topology also has another key advantage, higher efficiency due to decreased switching losses and also reduced output filter component size and cost as compared to a two level inverter.   Grid integration and disconnection: All grid-tied solar inverter systems are bound by certain regulatory standards.
Typically in the US, these standards are defined by bodies like the IEEE, UL (Underwriter Laboratory) and NEC (National Electrical Codes) NEC. The standards that govern grid-tied photovoltaic inverter systems are UL 1741 and IEEE 1547. One of the critical requirements of these standards is that all grid-tied inverters must disconnect from the grid if the ac line voltage or frequency goes above or below limits prescribed in the standard. See the table 1.7 below for such limits [18]. Also if the grid is not present due to a fault or for some other reasons, the inverter must shut down to avoid the solar power system of being an island. In any of these cases, the solar inverter system must not interconnect or feed power into the grid until the inverter is sure that proper utility voltage and frequency is recorded at the grid for a period of 5 minutes. This is to protect against the inverter feeding power into the grid during a fault or during a utility scheduled maintenance exercise. to drive the contactor that connects the inverter to the grid in the morning and remove it from the grid at night when the sun goes out.

The Key Components of a Grid-Tied Photovoltaic inverter system:
Looking at Figure   one junction more than the MOSFET, and this junction allows higher blocking voltage and conductivity modulation for reduced on-state conduction losses. The additional junction in the IGBT does however limit switching frequency during conduction.
MOSFET does not have this switching speed limitation. Inverters IGBT modules must be designed to be rugged, low loss and low on-state saturation voltages and must also be able to maintain a relatively high switching speed, up to 20KHZ with less loss.
A typical IGBT module and schematic from POWEREX [26] is shown below.   The main function of the transformer in a grid-tied solar power system is to convert the ac input voltage from the inverter core to the correct grid ac voltage both in magnitude and phase angle typically 208Vac RMS or 480V RMS Delta-Wye depending on configuration. The transformer also serves to provide isolation between the grid and the PV. The efficiency of the power magnetic (transformer and Inductor) is key to the efficiency of the solar power system. Some of the key specifications that define an efficient transformer are; very low No-load and Load losses, low winding temperature rise, good dielectric voltage withstand, good load regulation less than 1.5%, ability to withstand transient overvoltage situation up to 120% of rated voltage without saturation and no excessive audible noise (IEEE-1547). The efficiency of grid-tied inverter system is typically measured using the CEC (California Energy based on load levels and then the cumulative of all the load levels would be the overall efficiency. One of such transformer that I tested was a 100KVA 208/480V delta-wye transformer with a CEC efficiency spec on the name plate of 99.30% at 40 degree C ambient temperature, the physical picture is shown in Figure 1.34 below. This thesis has also presented details of the various regulatory standards that a gridtied solar power system must meet and the challenges involved in meeting these requirements. The thesis has shown that to achieve an efficient solar power system, it must start from the solar cell/module selection phase. Optimum mounting angle and the use of tracking where necessary to capture more sunlight is also an effective way to maximize efficiency. It also requires efficient photovoltaic inverter system with effective MPPT features. Key components of the inverter like the power switches, magnetic components must be properly selected for optimum performance. Lastly, this thesis has provided in detail the various ways to improve the efficiency of a solar power system from the solar cell structure to the array to the DC-AC inversion and grid interconnection. From the foregoing, improving solar power system efficiency involves careful consideration of all the various portion of the entire solar power system.