RET II Solar Cells - RET II Sommersemester 2020 PDF

Preview

Summary

Solar Cells - IntroductionThe sun was a mystery for our forefathers for a long time. It took a while to understand how to use solar radiation 1 and convert it into energy, to understand the sun’s mechanisms and fusion processes within the sun.The solar constant E 0 gives the amount of solar radiatio...

Description

TUM SOM

RET II

SS 20

Solar Cells - Introduction

The sun was a mystery for our forefathers for a long time. It took a while to understand how to use solar radiation1 and convert it into energy, to understand the sun’s mechanisms and fusion processes within the sun. The solar constant E0 gives the amount of solar radiation which is impinging 2 on the outer atmosphere on the surface of the size of 1m2, which is (exactly) perpendicular3 to the solar radiation. The solar constant can be measured outside the atmosphere perpendicular to the sun’s irradiation4 . è With E0 we can calculate the solar radiation, which is at the surface of the sun (ES) by putting two spheres into relation: o Surface area if the sun (rS2 * π) o We assume an imaginary sphere around the sun, where the earth’s center is on the sphere’s surface (rSE2 * π) è How much power in each second is radiated (ϕS) from the sun and the mass loss (m) can be calculated from E0 è Solar power on spherical atmosphere gives the amount of radiation which is coming on the earth’s surface: !" * E0 EØ = !Ø The calculation gives the amount of a sphere which is perpendicular to the solar radiation (half-sphere). Therefore, corrections/adjustments are needed. E0 is not a “real” constant, but it is varying in time. Although it is not a large variation, the scientific challenge lies in understanding these variations. In addition, climate change cannot be explained by the variations as many primarily thought. The Planck Radiation Formula calculates the spectrum of solar light, which comes to the earth. The Planck’s radiation law describes the wavelength of the emitted radiation, which is inversely proportional $ to its frequency (λ = ). This formula will determine the efficiency of semiconductors in solar cells. %

This graph of a spectrum is not shown in the conventional way (wavelengths and frequencies on the axes), but with energies (eV and MW/m2*eV) due to the fact, to describe mechanisms. We can observe that the overall radiated power (~ 63MW) can be recovered from the spectrum or calculated by intergrading the curve of the blackbody. A blackbody is an idealized physical body that absorbs all incident5 electromagnetic radiation, regardless of frequency or angle of incidence6. The spectrum of a blackbody, which is more shifted to higher energies/frequencies and lower wavelengths, depends on the temperature of the blackbody.

1 2 3 4 5 6

Strahlung Eindringen, aufschlagen shortest line that can be drawn between two parallels Einstrahlung einfallend Einfallswinkel

1

TUM SOM

RET II

SS 20

The sun’s position in the sky at a certain time of the day can be calculated from astronomy. Solar radiation interacts with the earth atmosphere. The incoming solar radiation7 (100%): • Gets reflected (~30%) through the atmosphere (especially on white clouds or snow-covered surfaces), back-scattered by the air (Rayleigh-scattering takes place on molecular air components, which are smaller than the wavelength of the lightàIncreases with decreasing wavelength) and reflected by the surface (Mie-scattering is due particles greater than the wavelength of the light. Strongly dependent on the locationàHigher in industrial areas, low in mountains). Or • gets absorbed (~20%) through clouds and the atmosphere (made by diverse gas components strongly dependent on solar spectrum. Mainly O3, O2, H2O, CO2). Solar light can also be measured, e.g. with a pyranometer (is a type of instrument for measuring the heating power of radiation, used for measuring solar irradiance on a planar surface and it is designed to measure the solar radiation flux density from the hemisphere) or measured /predicted by data sources derived from forecast methods. An example for data sources is MERRA. The Modern-Era Retrospective analysis for Research and Applications, provides data. The assimilation system enables assimilation of modern hyperspectral radiance and microwave observations, along with GPS-Radio Occultation datasets. It also uses NASA's ozone profile observations and takes some significant steps towards GMAO’s target of an Earth System reanalysis. MERRA-2 is the first long-term global reanalysis to assimilate space-based observations of aerosols and represent their interactions with other physical processes in the climate system. Forecast methods are quite important when it comes to installing PV-modules or the current electricity market prices. For example, power drops can have a major impact and cause big problems, because of solar eclipse or weather storms. Therefore, forecasts are required to predict such scenarios. Forecast methods can be differentiated in: • •

Short Term Forecast Methods (Astronomy/Weather Forecast and Data Driven Forecast; 1-2 day forecast) Long Term Forecast Methods (Astronomy/Climate Data combined with Building/Landscape/ Network Data) o Example: The LoD2 3D-Model (used in Greifswald) measures the global radiation (with consideration of hourly radiation and annual shadows on a specific roof) and calculates/ predicts the position and yield8 of the installed PV. Executive Summary:

We have lots of tools on how the solar light interacts with the environment and we can calculate the amount of radiation. Astronomy is a very strong point of solar radiation. Thanks to astronomy we at least know the max. radiation at each point in the world. Weather phenomena can alter solar radiation coming to the earth. Therefore, we need data/forecasts to get a better understanding of the real irradiation. Data can be derived from global data from past years/decades and from technologies which forecast irradiation for coming days.

7 8

Global radiation = direct radiation + diffuse radiation Produktion, Gewinn, Rendite

2

TUM SOM

RET II

SS 20

Solar Cells – Fundamentals

In order to understand the theoretical limit of the conversion of light into work, the principle of an ideal solar heat engine is required. The process can be divided in two parts:

1. Black absorber (lights comes from the sun and gets absorbed): any absorber who can absorb light, can also emit the light again (very basic effect). The amount of reemitted light strongly depends on the absorber’s temperature (TA). If the absorber has the same temperature as the sun’s surface, then the absorber will reemit the same amount of light, which has been absorbed (there wouldn’t be any gain/energy). Therefore, the absorber’s temperature must be lower as the sun’s temperature (TS). Only a small fraction will be emitted (it can also be calculated with Planck’s law of a blackbody). • η absorber = 1 – TA4/TS4 (ηa is very high, when TA is very low) 2. Carnot-engine: converts the light, which was converted to heat, into work. The conversion of heat into work is determined by the Carnot-efficiency. • η Carnot = 1 – T0/TA (ηC is very high, when TA is higher than the temperature of the environment, T0)

η total = η absorber * η Carnot è From this formula it can be seen, that is required to find an optimum for TA, hence there are two conflicting number for the absorber’s temperature! The max. efficiency of solar energy conversion theoretically could reach >80%, but current technologies are far away from it. Next step is to understand how semiconductors work, which are very popular when it comes to making PV cells. The Band Theory of Solids implies that the energy levels of sodium atoms become bands as their internuclear distance decreases. Putting many atoms close together, we get a huge band. In an atom, the electrons, which surround the nucleus of the atom, can only be very discrete energy states. These states can also be divided into other states due e.g. the angular momentum of the electron. If we have a solid a state, all this ladder of atoms spread to something, which is not a discrete number. We have two kind of bands: the valence-band and the conduction-band. Between these two bands there is also a third, the forbidden band. This band forms an energy gap. The size of the gap determines whether it is an isolator, conductor or a semiconductor. The basic process runs as follows: A photon hits an electron. Photons always have quantified energy, which is determined by wavelengths or frequencies of the photon (E = h*f). The photon’s energy is transferred to the electron. With the additional energy, the electron can be lifted into the conduction band and there remains a so-called “hole” in the valence-band. If the conduction-band is completely full, then there is NO conduction, meaning that the electrons cannot move up. Doping is a technique used to vary the number of electrons and holes in semiconductors. By bringing donor impurity 9 levels closer to the conduction band, the atoms can be easily excited to go into the conduction-band (not a good conductor, as there are more electrons in the band).

9

Fremdatom

3

TUM SOM

RET II

SS 20

The other way around would be introducing acceptor impurity levels: Electrons from the valence-band can move to the acceptor impurity levels. The missing electron in the valence-band is called “hole”, which then can transport electricity by moving through the band. Steps: 1. Charge generation (light impinges the semiconductor): an electron and a hole are generated 2. In the electron-hole-pair the photon’s energy is stored 3. These charges are separated by a p-n-junction (mainly through asymmetry) 4. Putting the recovered energy in an overall-grid to get the charges out of the semiconductor and to drive some load in the circuit Solar Cells – Efficiency

There is a natural limitation (e.g. absorption) of the conversion of the light into electricity of the solar cell. If a photon hits the PV-cell, where the energy is below the band gap, the electron cannot be lifted from the valence-band into the conduction-band (nothing happens), it just passes the cells. All photons below the band gap would be lost (1st limitation). Therefore, a low band gap would be perfect for getting a bigger portion out of the solar spectrum. If the photon energy is above the band gap, the electron would be lifted or be on a higher position in the conduction-band. Due physical processes or vibrations, the electron falls back down onto the edge of the conduction-band. The extra energy of the photon/electron gets lost and only energy related to the band gap will remain (only the band gap’s energy will be delivered by photons). A huge amount of energy will be lost again (2nd limitation). There are two extremes regarding the band gap: • High band gap: less photons are contributing but with more energy, but there is a huge loss • Low band gap: more photons with low energy are contributing to the spectrum, with small losses With optimization these extremes can be adjusted. The difficulties are in finding the right materials for the band gap. One possibility could be silicon solar cells. There are technical losses, which cause less efficiency, as well. Technical losses are categorized in: • Optical: o Reflection o Transmission o Shading

• Electrical: o Ohmic: § Bulk materials § Contacts o Recombination (hole-electron): § Surface § Bulk § Impurities

4

TUM SOM

RET II

SS 20

Possible ways to make PV cells with higher efficiencies are: • Tandem cells: The efficiency of two solar cells in tandem operation (at the top a high band gap and at the bottom a low band gap). à the photon’s energy won’t be completely lost. • Quantum Dot (is like the Tandem): dots absorb solar light, depending on thickness (just a theoretical idea). • Concentrator Cells (CPV)

Solar Cells – Technology

A

PV-device

works

as

follows:

1. Charge generation: Light excites electrons, freeing them from atomic bonds and allowing them to move around the crystal. 2. Charge separation: An electric field engineered into the material (pn-junction) sweeps out electrons. 3. Charge collection: Electrons deposit their energy in an external load, complete the circuit. The overall efficiency of converting light into work (electricity) can be >80%, but today’s devices are not even close of reaching this efficiency (there is lot of room for improvement). Solar cell types are classified into generations, which are equivalent to improvements or challenges in technology: • I Generation (Bulk Crystalline Silicon): Monocrystalline Silicon (sc-Si) and Multi-crystalline Silicon (mc-Si) are the dominant technologies and have the biggest market share. They are mature in efficiency (15-20 %) and structure (are very stable and are at the state of the art10 ). On the other side they are rigid / inflexible and require lot of energy in their production. • II Generation (Thin Film): are used for smaller devices (e.g. watches) and they are cheaper in construction (less material), but the cleaning process takes a lot of energy. Another disadvantage is low efficiency and life-time. E.g. Amorphous silicon (a-Si), thin film crystalline silicon (TF-Si). • III Generation (future approaches?): need much less material and are built for a very cheap price (e.g. printing technology). Again, life-time and efficiency are very low. E.g. Organic, quantum dot. Global market share and technology development show that the numbers are still increasing. PVtechnology can be applied in many applications and has made his way to the market (especially sc-Si and mc-Si) and will not be stopped again. For example, 2018 production numbers reported a total PV module production around 103 GWp. Nevertheless, the production of PV-modules is very cost and energy intensive, therefore it is necessary to look for (also more environmentally friendly) alternatives.

10

neuester Stand (in der Entwicklung von etwas)

5

TUM SOM

RET II

SS 20

Producing, cleaning and sawing the silicon is quite a dominating part of the costs. The challenge lies in reducing these production costs for silicon as far as possible and making the production process easier. How can we reduce the costs of PV-modules further? Just the reduction of module-costs won’t be enough, also the reduction of BOS 11-costs must be considered. It is necessary to make a balance of plants to be more sophisticated 12 but also cheap at the same time. It’s not enough having a PV-module, which is just a passive element in the grid dominated by a huge power plant, but a module forming an entity 13 within a smart grid (is quite a challenge!). Silicon is a very abundant material, it is making 27.7% of the Earth crust by mass. In nature silicon dioxide SiO2 (Silica) is found as sand, quartz. It is non-poisonous and there are no problems with the waste. c-Si based Solar Cell - Production Flow and Technologies - From raw material to module: à Metallurgical Grade Silicon still has too many impurities/recombination, which cause low efficiency à The Siemens-process (purification process) eliminates the remaining impurities à Besides solar grade silicon, electrical grade silicon can be made, which is found in electrical devices (but the production is more expensive) à Multi-wire inner diameter slicing is a process which has a lot of losses

The Metallurgical Grade Silicon (MG Si) Production is a process of converting silica into silicon, more precisely it’s a carbothermic reduction (SiO2(s) + 2C(s) = Si(l) + 2CO(g)): Silicon dioxide and carbon are mixed together, and electrodes heat the materials up in a crater. The liquified metal undergoes a solidification (reduces impurities) and is then crushed and sized into silicon. The recovered gas (CO) from the heating process undergoes a further reduction and is converted into CO2, also energy and traces of silica are recovered. The energy demand from SiO2 to solargraded Silicon (SG) is 176 kWh/kg: • In order to build a 100 Watt PV-module, ~ 0.7kg SG-Si are needed • To create 1kg SG-Si14 by the Siemens Process 4.4kg MG Si are needed • 19kg CO2 are emitted through that process

11 12 13 14

BOS (Balance of System) includes all further parts other than the modules (cables, mounting equipment...) fortgeschritten Einheit SG Si: 99.99999999% purity

6

TUM SOM

RET II

SS 20

Ingot growth – From SG-Si to Monocrystalline ingot (Czochralski (CZ) process): • • • •

Standard for integrated circuit industry High quality ingots Growth rate approx. 5 cm/h Disadvantage: Quartz crucible causes oxygen and carbon impurities of the melt.

Typical crystal size: Diameter 10-30 cm Length 1-2 m

Ingot growth – From SG Si to Monocrystalline Ingot (Float zone (FZ) process): • No crucible – highest quality! • Higher price.

From SG-Si to Multi-crystalline ingot

7

TUM SOM

RET II

SS 20

Now we have the semiconductor layers for a solar cell. The grain size in multicrystalline silicon is from several microns to several millimeters or even centimeters. The fundamental physical properties such as bandgap and absorption properties are similar. The difference between sc-Si and mc-Si is primarily the density of defects and impurities and COSTS. In addition, mcSi have higher probabilities for recombination, depending of course of the size of grain boundaries.

Next step would be bringing the ingot into a wafer (wafering). Wafer thickness determines a cell’s absorption (if thickness is reduced, so is the absorption). There exist two sawing processes: Wiresawing and inner diameter slicing. Both techniques have high kerf losses (30-40%), which is a overall problem. A possibility to reduce kerf losses as much as possible is the so-called Ribbon Growth on Substrate (RGS) – wafer manufacturing. This process has a larger throughput/day and it reduces material consumption by producing thinner wafers and has no kerf losses. Downside once again, is a lower efficiency of solar cells with ribbon cut wafers. RGS Principle: • Substrate below melting point of silicon (1414° C) is moving under a frame with liquid silicon. • The substrate is extracting heat and forces a crystallization process of the melted silicon starts. • By leaving the frame, crystallization process is stopped. • Due to the different expansion coefficients the wafer will separate from the substrate Wafer to Solar Cell

à

Cells to Photovoltaic Modules

8

TUM SOM

RET II

SS 20

Solar Cell – Application

First discussion to start with, regarding the application of solar cells, is about potentials. Whenever we deal with RET, first question that comes up, is about the potential of the technology. Solar cells might be THE technology which can be used more or less everywhere and if the production of cells is under control for once, it also brings environmental advantages (other sources have limitations, e.g. hydropower, wind-power or biomass15). Island systems started the whole PV-deployment. People thought at the very beginning that even when systems very expensive, island systems might be a very good choice for remote areas, compared to e.g. diesel. Still in many parts around the world, if there is no access to a main grid, people usually use diesel-generators to produce electricity. In the future, more PV-modules will be applied. The electricity price of solar cells is much lower (~ factor of 6) than diesel (compared to a coal power plant it is still a low price), but a storage (e.g. battery) is needed. Diesel generator could still be built as a backup and the solar panel would work as a fuel saver. RET always tries to find a system (in an hourly resolution), which can supply the energy for a whole year. The model is very simple: It should always be adapted to the electricity load. Depending on costs and ability of the PV-module and battery, a diesel generator might be needed as ba...