Repositorio Institucional de Documentos

Analysis of aerosol particles for the determination of energy losses in solar power plants Thesis of: Juan Prats Briceño Study course: Mechanical Engineering Major topic: Solar energy Saragossa, October 2012 This thesis was supervised by: Mr. Dr. rer. nat. Markus Sauerborn Solar-Institut Jülich Fachhochschule Aachen Mr. Prof. Dr.-Ing. Cristobal Cortés Energy department Universidad de Zaragoza This thesis was written independently. There was no other literature used except the mentioned ones. Saragossa, the 02.10.2012 Acknowledgments This thesis was done at the Solar-Institut Jülich (SIJ) of the Aachen University of Applied Sciences, Division Jülich. Preliminary, I would like to thank people who helped me on my way to finish my work. Markus Sauerborn showed the way to write my thesis. He explained me really good what was happening with the solar radiation and the aerosols, and he showed me how the devices worked. He was always there when I had a problem. Kai Barkschat helped me to solve the initial problems that I had with the ceilometer. I am really thankful to Max Wagner, he was always interested in my thesis and he gave me good tips. His help was also useful at the time to use the camera system and the software Visual Builder. My girlfriend Eva helped me very much. I had some problems to write the thesis in a foreign language, but she helped me with each doubt. She also gave me good tips in order to make a better structure of the thesis. Abstract This thesis is a study about the influence of the aerosol particles on the circumsolar radiation. The research started in October 2011 with the reading and learning about other studies on the CSR and aerosol particles. It was also necessary study the functioning of the devices, which were going to be used. For each machine it took about one week to understand how it worked and they were then tested one by one. In December 2011, once some initial problems had been solved, the measurements could finally start to be taken, and by the end of January all of the data was recorded. The calculations and the literature work has been done in the city of Saragossa, Spain, since February 2012 until October 2012. Table of contents 1 Introduction and goals .................................................................................................... 7 2 Solar radiation ................................................................................................................ 9 2.1 Solar energy ................................................................................................................. 9 2.2 Spectral distribution of the solar radiation ...................................................... ..... …10 2.3 Losses in the atmosphere ........................................................................................... 11 3 Circumsolar radiation ................................................................................................... 14 3.1 Description ................................................................................................................. 14 3.2 Circumsolar ratio ............................................................................................. ..... …15 3.3 Previous studies ......................................................................................................... 15 4 Aerosols ........................................................................................................................ 19 4.1 Description ................................................................................................................. 19 4.2 Composition .................................................................................................... ..... …19 4.3 Origin… ..................................................................................................................... 19 4.3.1 Of natural origin .................................................................................................. 20 4.3.2 Of anthropogenic origin ...................................................................................... 21 4.4 The influence of the aerosols on the solar light scattering ........................................ 22 4.4.1 Direct effect ......................................................................................................... 22 4.4.2 Indirect effect....................................................................................................... 23 4.4.3 Impact of direct and indirect aerosol effect ......................................................... 24 5 Clouds ........................................................................................................................... 25 5.1 Description ................................................................................................................. 25 5.2 Reflection caused by the clouds ................................................................................ 25 5.3 Types of clouds… ...................................................................................................... 28 5.3.1 High clouds .......................................................................................................... 29 5.3.2 Middle clouds ...................................................................................................... 30 5.3.3 Low clouds .......................................................................................................... 31 6 Ceilometer: CHM15K .................................................................................................. 31 7 Aerosol spectrometer: FIDAS 200 ............................................................................... 38 8 Weather Station ............................................................................................................ 41 8.1 Instruments… ............................................................................................................ 41 8.1.1 Weather mast ....................................................................................................... 41 Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 6 8.1.2 Solar tracker......................................................................................................... 41 8.2 Measurements of the solar radiation… ...................................................................... 42 8.2.1 Global and diffuse radiation ................................................................................ 42 8.2.2 Direct radiation .................................................................................................... 43 8.3 Other measurements… .............................................................................................. 44 8.3.1 Measurements of the air temperature .................................................................. 44 8.3.2 Measurements of the relative humidity ............................................................... 44 8.3.3 Measurements of the air pressure ........................................................................ 45 8.3.4 Measurements of the wind speed and direction .................................................. 45 8.3.5 Measurements of the precipitation ...................................................................... 45 9 Camera system.............................................................................................................. 46 9.1 Telephoto lens ............................................................................................................ 48 9.2 CCD sensor ................................................................................................................ 48 10 Measurements ............................................................................................................... 51 10.1 Results of the measurements ..................................................................................... 51 10.1.1 14th of December 2011 at 13:31h ........................................................................ 51 10.1.2 10th of January 2012 at 14:03h ............................................................................ 54 10.1.3 11th of January 2012 at 11:34h ............................................................................ 54 10.1.4 13th of January 2012 at 11:06h and 12:17h ......................................................... 60 10.1.2 16th of January 2012 at 12:15h, 12:45h and 14:45h ............................................ 62 10.2 Relation between the measurements .......................................................................... 65 11 Conclusion .................................................................................................................... 74 Bibliography ........................................................................................................................ 77 A1 Calculations .................................................................................................................. 79 Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 7 1 Introduction and goals This thesis was written for the “Solar-Institut Jülich” (SIJ), located in Jülich, Germany. The SIJ was founded in 1992 as a scientific institution of the university of applied sciences “FH Aachen”. The aim of the institute is the research and development in the field of renewable energy, especially in solar energy. One of the most important projects in which the SIJ participates, is the “Solarturm Jülich” (STJ). The STJ is a high temperature solar thermal power plant. It has an absorber on the top of the tower, which receives the solar radiation reflected by 2150 heliostat. The absorber is made of ceramic and it transmits the solar radiation to the working fluid, air in this case, to a temperature of 700ºC. This air heats steam and it works in a steam turbine to generate as maximum 1500 KW. /1/ This kind of high temperature solar thermal power plant can have an efficiency of about 17%. Losses that are normally considered are, for example, due to the wind, cosines losses, reflexion or absorption /2/. But there also are losses, which are disregarded and this thesis is going to be focused one of these losses. The aerosol particles in the air of the atmosphere can scatter the solar radiation, deviating its direction slightly. This causes what is known as circumsolar radiation, which is still measured as direct radiation (the incoming energy in a high temperature solar thermal power plant), but it reaches the absorber offering a more divergent beam with less concentration. For this thesis, four different devices have been used: a ceilometer, an aerosol spectrometer, a weather station and a CCD camera. The ceilometer is a meteorologic instrument using reflected laser light to measure cloud and the particles layers in the atmosphere from the ground level up to a height of 15000 m. With the aerosol spectrometer the concentration and size of aerosol particles on the ground level can be measured. The weather station of the SIJ is made up of many different devices. The most important ones being the pyrheliometer and the pyranometer, which are used to measure the direct radiation and the global radiation. Finally, a CCD camera with a higher resolution (5 megapixel) was used: it was connected to a computer, where the pictures were recorded and analysed by a program. A telephoto lens, a filter and an automatic tripod were used to take the photos. The pictures were then measured with software in order to calculate the radiation coming from the circumsolar area. These measurements were taken on five different days, in eight different moments. Once these measurements had been taken, they were compared and analysed in order to see whether there was any influence of the aerosol particles on the circumsolar radiation or not. The figure 1 shows the place where the measurements were Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 8 taken. The majority of the devices were located on the flat roof of the “Naturwissenschaft” building of the SIJ (Jülich, Germany). One can see the camera with the telephoto lens and the linked computer in the foreground. The ceilometer is located in the background on the left and the aerosol spectrometer next to it. The pyrheliometer and the pyranometer cannot be seen in the photo, but they were just a few meters up fixed on a solar tracker. Figure 1: Place on the institute flat roof where the measurements were taken, in Jülich, Germany The thesis is divided in 4 big parts and in 9 different chapters. The first part explains the necessary theoretical details in order to understand the thesis. It includes the chapters Solar Radiation, CSR, Aerosol Particles and Clouds. The second part describes how the different devices used in this thesis work and also specifies their technical characteristics. It is divided up in the chapters CHM 15K, Fidas 200, Weather Station and Camera. The third part includes the chapter concerning the measurements, in which all of the results are first listed and then compared. The last part consists of the appendix, in which the calculations are shown in detail. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 9 2 Solar Radiation 2.1 Solar energy: The sun is a huge sphere composed of high temperature gases. The diameter of the sun is 1.39x10 9 m /3/ and it is situated 1.49588707x10 11 m from the earth; this distance is called Astronomical Unit (AU). /4/ The sun generates its energy by nuclear fusion reactions, which are produced in the nucleus. So the energy come from the mass loss of the sun, given by the equation E = m c2. The solar nucleus covers 15% of the volume of the entire sun, but it makes up about 40% of the total weight and generates 90% of the energy. The density in the nucleus is approximately 10 5 kg/m3 and its temperature can reach 10 7 K. The photosphere, which is the outer layer of the sun, has a radius making up approximately 30% of the entire sun’s radius. It is estimated that the temperature in the photosphere is the “effective temperature of the black body”, which is 5762 K, and the density is below 10 5 kg /m3. The photosphere is considered the surface of the sun because it is an opaque region, where the largest part of radiation is emitted to space. /5/ Three factors determine the luminous flux coming from the sun towards the earth: the distance between the sun and the earth, the solar diameter and the temperature of the sun. The flux emitted by the sun is in engineering usually considered as a constant. Meteorological changes influence the flux in much greater ways than the variation in solar radiation coursed by the changing distance to the sun on the elliptic way around the sun. The solar constant (G) is the value of the energy flux coming from the sun. A flux is something that passes through a surface, in this case, energy per second. The unit W/m2 is thus used. G measures the value of the perpendicular solar incident flux to a surface, G is valid out of earth’s atmosphere, when the solar radiation reaches the atmosphere it suffers changes. The energy flux emitted by a black body (surface of the sun) is given by the Stefan- Boltzmann law. This law states that the energy radiated by a black body is directly proportional to the fourth power of the black body’s temperature. In order to calculate the solar constant, the energy flux emitted by the sun is multiplied by the relation of the areas between the sun’s surface (rs) and a sphere located at one astronomical unit (ao) to the sun. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 16 Aerospace Centre (Deutsches Zentrum für Luft- und Raumfahrt, DLR) in Cologne, Germany. The LBL collected data for about ten years from eleven different places across the United States. The locations have distinct atmospheric characteristics (for example humidity, altitude, climate, proximity to sources of large particles, etc.), and a wide range of data was obtained. About 200000 measurements were taken from the mid to late 1970s and in the early 1980s. The measurements were all taken with the same telescopes, which measured a radial profile of the sun out to an angular displacement of 56 mrad. The telescope used as its basic optical element an off-axis mirror of 7.5 cm diameter and 1 m focal length. The telescopes functioned automatically and on its own for up to one week and collected one measurement every minute. A fused silica window protected the mirror from the environment. The mirror formed an image of the sun and the sky around it on a plate next to the telescope axis. A small hole in this plate, the detector aperture, defined the angular resolution (1/20 of the solar diameter), and the amount of light passing through the aperture into the detector assembly constituted the fundamental measurement. In the detector assembly the light was mechanically chopped, optically filtered, and focused onto a pyroelectric (thermal) detector. This type of detector was chosen for its uniform wavelength response in the 0.3 to 2.5 micrometer region, as well as its wide dynamic range. The telescope scanned through a 6 degree arc with the sun at the center and measured the intensity of the solar and circumsolar radiation based on the angle. The instrument scanned in declination so that at sunrise and sunset it travelled nearly parallel to the horizon and at noon it moved in a vertical plane. Each 6 degree scan required 1 minute of time. The intensity was registered every 1.5' of arc. Within 0.5 degrees on either side of the sun, an aperture of a size of 1.5' of arc was used, and outside this region the aperture was increased to 4.5' of arc. One set of measurements consisted in one scan at each of 10 filter positions: eight optical filters, one open (or clear) position, and one opaque position. The opaque position was used to measure the detector noise. The absolute determination of the normal incident flux (within 2.5 degrees of the sun's center) was provided by an active cavity radiometer. This device was selfcalibrating and had an accuracy of 0.5%. The pyrheliometer was provided with a matched set of filters, which rotated synchronously to those on the scanning telescope. Thus the telescopes produced an absolute measurement of the normal incident flux, along with the Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 17 detailed solar profile in eight wavelength bands. Two pyranometers were used, one set up in the usual horizontal position and the other one tracking the sun. The results show that there was a big variation of the CSR data depending on the place where the measurements were taken. These locations were very far apart and they had very different conditions, but these conditions were not considered in order to compare the obtained CSR data. In the figure 6, one can observe the fluctuations in the measurements. /12/ The DLR have designed and created a digital sunshape camera using an optical telescope together with a 12-bit digital resolution CCD camera. Using this camera, 2300 solar profiles were acquired from three sites across Europe: the DLR site in Cologne (Germany), the PSA in Almería (Spain) and the CNRS Solar Furnace in Odeillo (France). These profiles were grouped and averaged according to each sunshape’s CSR, using a method similar to the one applied to the RDB in this thesis, however, accepting larger ranges of CSRs for each sunshape bin. The optical system of the DLR’s telescope includes a band pass, neutral density filter. This is used together with the CCD camera in order to improve the spectral response of the CCD’s silicon wafer. The solid angle subtended by each element in the CCD camera used by the DLR telescope is 0.065 millisteradians (msr). The equivalent figure for the LBL’s telescopes was 0.44 msr, which demonstrates the lower spatial resolution of the LBL data. This improvement by the DLR provides both a greater number of data points to be recorded across the transition between the solar disk and the circumsolar region, and higher quality data for each image point. Also, the time required for the DLR to acquire an image with the CCD camera was virtually instantaneous, compared to scans from LBL lasting 1 min. The shorter time span provides a higher probability of acquiring self-consistent profiles. /11/ In the figure 6, one can observe and compare measurements taken by the DLR and by the LBL. Each colour represents a measurement of the brightness of the sun sphere considering its angular displacement. One can observe that these measurements had different results. /12/ Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 18 Figure 6: results of the measurements made by the LBL and the DLR /12/ Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 19 4. Aerosols 4.1. Description Technically, an aerosol is a suspension of fine solid particles or liquid droplets in a gas. The sizes of these particles can vary from 0,002 mm to 100 mm. The smallest ones are molecules in the air, which can remain in the air or a gas for weeks, but the larger they are, less time they can remain suspended. Particles with a size over 10 mm cannot be suspended more than a few hours due to the gravity effect. The majority of the small particles disappear because of precipitations. Particulate matter (PM) is the term used for liquid droplets or solid particles suspended in a gas. There are two main groups to distinguish the PM: coarse and fine particles. Coarse particles have an aerodynamic diameter larger than 2.5 mm, and fine particles one smaller than 2.5 mm. The aerodynamic diameter of an irregular shaped particle is the diameter of the spherical particle with a density of 1000 kg/m³ that has the same settling velocity as the irregular particle. In this thesis, the sizes which will be used are PM 1, PM 2.5, PM 4, PM 10 and PM total. They refer to particles with an aerodynamic diameter smaller than 1, 2.5, 4, 10 mm and the total amount of the particles suspended in the air. Aerosols can also be measured by weight per volume. The total amount of the aerosols in the air varies between 1 mg/m³ (as can be found at the polar cap or in the free troposphere above the oceans) and 1000 mg/m³ (in a wildfire or in desert sandstorms). 4.2. Composition: The aerosols can also be classified in different groups according to the substances they are made up of, such as sulphates, organic carbon, black carbon, nitrates, mineral dust, and sea salt. In reality, however, the particles are usually more complex, consisting of more than only one substance, which makes it difficult to group them according to their composition. /13/ 4.3. Origin The majority of the aerosols (about 90 percent) is of natural origin, like volcanic ash, the smoke of wildfires or the mineral dust in deserts. The remaining 10 percent has anthropo- Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 20 genic origin, mainly from the burning of fuel. While these aerosols appear much less, they can be found at very high concentrations in the air in big cities. The most common aerosols are: 4.3.1. Of natural origin: Soil dust: it comes from elements which occur in the composition of soils. Due to the effect of wind this mineral dust is eroded from the crust. These aerosols consist mainly in oxides (SiO2, Al2O3, FeO, Fe2O3, CaO, etc.) and carbonates (CaCO·, MgCO3) that which appear in the earth’s crust. The annual mineral dust emissions are estimated between 1000 and 5000 millions tonnes per year, the majority of which comes from the deserts. The main sources are the Sahara and the Gobi desert. The figure 7 shows the dust concentration on a global scale. Although the source of these aerosols is considered natural, it is estimated that around 30 percent of the mineral dust is produced due to the human action, by activities that increase the desertification. Figure 7: dust concentration in the world /13/ Sea salt: when a wave in the sea breaks, because of the wind, some amount of water gets into the air in the form of small particles. The composition of these particles is the same as that of seawater: substances such as water, sodium chloride and magnesium and sulphate salts. This type of aerosol scatters the sun light and forms clouds. The sea salt particles are quite large, so they cannot travel very far away and usually fall down over the sea. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 21 Primary biogenic aerosols: they are composed of plant debris, humic matter and microbial particles. They are present in areas with intense vegetation. Volcanoes: they eject huge columns of ash, sulphur and other gases to the air. The sulphate aerosol particles, which are in the upper troposphere, contribute to the formation of clouds. This contributes to the cooling effect called the “indirect aerosol effect”, further explained at the end of this chapter. The figure 8 shows measurements taken by the Mauna Loa Observatory, where the solar radiation which has been transmitted over the last fifty years is shown. The influence of the volcanic eruptions on the percentage of transmitted solar radiation is clearly visible. Figure 8: Solar Radiation Transmitted by Mauna Loa Observatory /14/ 4.3.2. Of anthropogenic origin: Sulphates and nitrates: these are secondary particles which are produced from the reaction of primary gases (sulphur oxide and nitrate oxide NOx) in acids (gaseous nitric acid and liquid sulphuric acid). They can be formed with or without ammoniac. The origins might vary: either of natural origin (volcanoes or sea plankton) or of anthropogenic origin (fuel combustion). Both types cause heavy light scattering. Organic and black carbon: Organic elements represent the largest single component of biomass burning aerosols. Organic aerosols are important constituents off the highest part Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 22 of the troposphere. They are water-soluble and contribute to the nucleation of cloud droplets. The black carbon has a light absorbing character. Industrial dust: coal combustion, traffic, construction, waste incineration and other industrial activities produce primary aerosol particles. In the developed countries, the amount of these particulates has decreased in the last years, but it is increasing in the newly industrialising countries. While this signifies a big environmental problem, it doesn’t have great consequences on the light scattering due to the large size of the particulates. 4.4. The influence of the aerosols on the solar light scattering In this thesis the influence of the aerosols on the solar radiation is studied, so it is necessary to explain how these particles affect the incoming solar radiation. Some first studies have been made about the capacity of the aerosols to modify the radiation balance. These studies were done with anthropogenic particles, like sulphate aerosol (SO2) or products from the burning of biomass. Scientists as Charlson Kiehl and Brieglev estimate that these aerosols could backscatter a quantity from -1 up to -2W/m2 to space. The aerosols can affect the climate in two different ways: directly and indirectly. 4.4.1. Direct effect Aerosol particles in the atmosphere reflect or scatter solar radiation depending on their size. As explained in Chapter 2, the solar radiation spectrum has a wide range of wavelengths, and the aerosols particles interact with those of a similar size. In the figure 9, four different examples of the interaction of particles and solar light are given. The first one shows the effect of a very small particle: there is not a big influence of the particle on the wave. When the size of the particle is very large (>10 mm), there is also no scatter. In the second example, large particles (over 2 mm) were used, and they caused Mie scattering, where the light is scattered at a small extent. The third one shows how Rayleigh and Mie scattering occur. The particles interact with wavelengths of a similar size, in this case with the particles in the range of the visible light, which contains the majority of the solar energy. The wave is reflected by the particle and it is scattered, part of it out to space. Furtheron, there are also particles that absorb light. They are coloured, and the radiation causes them to heat up. The large particles previously mentioned can also be coloured and absorb the sunlight. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 23 Figure 9: Interaction of solar light and particles /15/ 4.4.2. Indirect effect Cloud formation is dependent on aerosols. If there are not aerosols, super saturation (relative humidity) can be seen without droplet formation. Cloud droplets need aerosols to be condensed. If not much particles are present (less than 200 per m³) the droplets will be large, but if there are a lot of aerosols present, smaller droplets will be formed. Clouds with big droplets reflect less light compared to clouds with small droplets. The high cirrus clouds, which have large droplets, do not reflect much solar light but they reflect the long wave infrared radiation better. So this way aerosols have an indirect effect on the climate. This will be explained more in detail in Chapter 5. The figure 10 shows the relation between the number of aerosols particles and the number of droplets. As can be seen, the relation is not lineal, as of 1000 particles per m³ the effect stops increasing and remains at a steady constant. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 24 Figure 10. Number of droplets per volume in relation to the number of particles per volume. /15/ 4.4.3. Impact of direct and indirect aerosol effect: As estimated by the Intergovernmental Panel on Climate Change (IPCC), aerosols represent about 30% of the driving power of greenhouse gases. According to the IPCC, the warming effect of the greenhouse gases is 2.5 W/m², while the cooling effect of aerosols would be 0.7 W/m². After adding them, the net result is thus a warming effect of 1.8 W/m². Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 25 5. Clouds 5.1. Description The clouds have a very important impact on the scattering of the solar radiation. In this chapter, it is going to be exposed how solar radiation and the different types of clouds interact. A cloud is visible mass made up of liquid water droplets or ice crystals and other chemicals suspended in the atmosphere. Clouds are also considered as aerosols. The size of the diameter of these droplets varies between 5 and 100 m. In order for them to turn into a raindrop, their diameter needs to be bigger than 1 mm. The concentration of the droplets in a cloud varies between 25000 and 1 million droplets per air litre. Clouds significantly influence the climate and the solar radiation, both the incoming radiation from the sun and the outgoing radiation from the earth. The thick clouds, which are situated at a lower height, reflect solar radiation, thus cooling the surface of the Earth. The thin clouds, which are situated at a higher level, primarily transmit incoming solar radiation, but they also reflect the outgoing infrared radiation emitted by the surface of the Earth, which causes a warming of the Earth’s surface. /16/ 5.2. Reflection caused by the clouds Every surface reflects the light, and the clouds are not an exception. This reflection is called albedo. A high albedo means that a large portion of the sun’s energy is reflected; snow, deserts or clouds are examples of this. Then there are also low albedos, which only reflect a small part of the radiation and absorb the rest; oceans or rain forest surfaces have low albedos. About 30 percent of the incoming solar energy to the Earth is reflected back to space, mainly due to the clouds. In the figure 11 appears the different albedos for different surfaces and clouds. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 32 6. Ceilometer: CHM 15K For the measurements of the height of the aerosol particles and the clouds, as well as their concentrations, a CHM15k ceilometer was used. With this machine it is possible to make measurements of particles in a height of up to 15 km. The machine used in the work for this thesis is located on the roof of the Naturwissenschaft-building, on the FH Aachen campus, in the city of Jülich. The machine works based on the LIDAR-method (Light Detection And Ranging), an optical sensing technology that can measure the distance, and other properties of objects in the atmosphere using pulses from a laser. The method is similar to that of radar, which determines how far the target is away, depending on the delay between the emitted and the reflected signal. But in this case it works with light, not with radio waves. The LIDAR technology is mostly used to measure geographical characteristics and to develop maps, but it has more applications. LIDAR uses wavelengths between 10 micrometers and 250 nanometres. As a result, the spectrum reaches from ultraviolet to visible and infrared light. As discussed in previous chapters, the laser works in a range similar to that of the solar spectrum, so it is going to be useful to measure particles in the atmosphere, which influence the sunlight. These are clouds, raindrops, aerosols and even molecules. The laser pulses interact with the different particles via scattering. The most usual scattering mechanism is Mie scattering. Figure 19 gives a sketch of how the LIDAR technique works: Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 33 Figure 19: Operating principle of LIDAR. /20/ In the ceilometer, a laser points vertical towards the sky and there is a receiver next to it. The duration of the laser pulse is several nanoseconds, and it is directed vertically through the atmosphere. When the beam crosses through the atmosphere, a fraction of it is scattered by particles in the air, normally with sizes similar to those of the wavelengths of the laser. This form of scattering is called Mie scattering. A small part of the scattered light will return to the LIDAR receiver. The time t between the emitted laser signal and the signal measured in the receiver will give the distance of the particles, which were involved in the scattering process. The distance R can be calculated with the following equation: Equation 6: Distance /21/ Where cair is the speed of light in air. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 34 To measure the power of the light scattered back and received by the LIDAR detector, the following equation is used: Equation 7: Power of the light received by the LIDAR /21/ It consists of four different parts, which are described below: K is an experimental regulation parameter, which is calculated using the following formula: Equation 8: LIDAR experimental regulation parameter /21/ Where P0 is the average power of a single laser pulse, c is the speed of light in air,  is the time duration of a pulse, A is the area of the receiver and  is the total system efficiency. K is a system-specific parameter and constant in the system. G (R) is a geometry factor. It is calculated with the following formula: Equation 9: geometry factor /21/ Where O (R) is the overlap function of the laser beam and the field of view of the receiver. And R is the distance of the particles as mentioned before. The parameter  is the backscattering coefficient, which determines the power of the signal received on the ground. The equation is an addition of the intensity due to both Rayleigh scattering and the Mie scattering. Equation 10 backscattering coefficient /21/ T is the transmission term. It describes the quantity of light lost in the atmosphere before being scattered. With the Lambert-Beer-Bouguer law, these losses can be determined: Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 35 Equation 11 /21/ Where (R,) gives a ratio of how much intensity is lost after the light crosses through a distance R in a gas. The integral describes the path of the light from the laser and back to the receiver. The dependence of (R,) and (R,) of the wavelength of the laser light is determined by the size, refractive index and form of the scattering particles. In the work for this thesis a CHM 15k ceilometer from the German company Jenoptik was used, as shown in figure 20. Furthermore, the technical specifications of the ceilometer are given in table 1. Figure 20: CHM 15K used in this thesis. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 36 MEASURING PARAMETERS Measuring principle Optical (LIDAR) Measuring range 15- 15000m Accuracy ±5m Resolution of backscatter data Standard: 5m Optional: 15m Hardware resolution 200 Mhz (sampling rate) Time to measure 5s to 60min (programmable) Targets Aerosols, clouds Quantities to be measured - Cloud base (max. 5 layers, preset: 3 layers) - Cloud amount - Penetration depth - Vertical visibility - Height of mixing layer Light source Nd:YAGsolid - statelaser, wavelength 1064 nm INTERFACES AND SOFTWARE FOR DATA OUTPUT AND DEVICE CONFIGURATION Standard interface RS485 Optional interfaces RS232, RS422, LAN, CIBUS Communication Measured data and settings are transmitted in data telegrams. Easy device configuration and firmware upgrades with JO-Data Client software. Optional software JO-Visual Software for convenient visualizing measured results ELECTRICAL PARAMETERS Power supply 230 / 110 V(AC), ±10 % Power consumption 250 W (standard) 800 W (in maximum heating mode) OPERATING SAFETY Environmental compliance ISO 10109-11 Laser protection class 1M according to DIN EN 60825-1 Internal protection class IP 65 EMC Class B, DIN EN 61326-1 Electrical safety DIN EN 61010-1 Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 37 Certifications CE OPERATING CONDITIONS Temperature -40 °C ... +55 °C Relative humidity 0 % ... 100 % Table 1: technical specifications of CHM 15K /22/ Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 38 7. Aerosol spectrometer: FIDAS ® 200: For taking measurements of the aerosols on the ground level it has been used a light scattering photometer, the Fidas ® 200. The machine used in this thesis is located on the roof of the Naturwissenschaft building, in the FH Aachen campus in the city of Jülich. The Fidas® continuous ambient air quality monitoring service provides continuous and simultaneous PM1, PM2.5, PM4, PM10, TSP (total suspended matter) and the particle number concentration. The system uses the approved measurement technology of optical light scattering and is equipped with a LED light source with stable output and long lifetime. The machine operates with an aerosol flow of 5l/min and is equipped with a Sigma-2 sampling head according to VDI 2119-4, which allows a representative measurement even at strong winds. Fidas ® 200 provides an Intelligent Aerosol Drying System (IADS) as well as sensors for the measurement of ambient temperature, air pressure and relative humidity. Principle of operation: In the following scheme it can be seen the different parts of the Fidas ® 200, then is going to be explained its principle of operation. Figure 21: principle of operation of the Fidas 200. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 39 The air get into the sigma-2 sampling head, then the IADS avoids, that the particle size measurement is effected because of moisture, by using a dynamically regulated drying system in regards to relative humidity and ambient temperature. The aerosol sensor is an optical aerosol spectrometer that determines the particle size and number by a scattered light analysis after Mie scattering. The particle move separately through an optically differentiated measurement volume, homogeneously illuminated with white light. Each particle generates a scattered light impulse, detected at an angle of 85º to 95º. The number concentration is deducted from the number of scattered light impulses. The intensity of the scattered light is a measure for the particle size diameter. The lower detection limit was reduced to 180 nm by using optimised optics, higher light density and improved signal analysis (logarithmic analog digital converter). Therefore smaller particles, measured roadside in high concentration, are better accounted for. The better the classification precision and the resolution capacity, the much preciser is the definition of the particle size distribution. Using a white light source, a precise calibration curve without ambiguity can be achieved, resulting in an extremely high size resolution. The patented T-aperture leads to an accurately defined optical measurement volume and permits a particle measurement without border-zone-errors and therefore a precise size measurement. The new and quick digitalised signal electronic analysis allows the identification and correction of coincidence. For the conversion of the measured indicators into mass or mass fractions, the high resolution particle size distribution will be multiplied in each value with a correlation factor, representing that the aerosol particulate is build up from different sources (e.g. combustion aerosols, tire abrasion, pollen) according to its particle size. A mass fraction is achieved by applying an additional separation curve (e. g. DIN EN 481) to the determined particle size distribution. Downstream to the optical sensor there is a filter holder for an optional gravimetric validation of measured data. /23/ In the following pictures one can observe the machine Fidas ® 200, and bellow there is a table with its technical parameters. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 40 Figures 22 and 23: Fidas ® 200 Measuring principle Optical light scattering Reported Data (simultaneous) PM1, PM2,5, PM4, PM10, TSP, number Size channels (optional) 64 Measurement range (particle size) 0,18 – 18 μm Measurement range (number) 1 – 20000 particle/cm3 Measurement range (mass) 0 – 1500 μg/m3 Time resolution 1 s – 24 h Aerosol flow 5 l/min (0,3 m3/h) Working temperature -30 to +35°C Power supply 115/230 V; 50/60 Hz Power consumption 140 W Dimensions 18.5x45x32cm Weight 9.3 kg Interface Touch display 800 x 480 pixels Data logger 4 GB Compact Flash Network LAN, WiFi Table 2: technical specifications of Fidas 200 /23/ Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 41 8. WEATHER STATION: The Solar-Institut Jülich, where the measurements for this thesis were done, also features a weather station. This station was used to determine the following meteorological parameters: temperature (ºC), relative humidity (%), wind speed (m/s), wind direction (º), air pressure (hPa), precipitations (mm) and the global direct and indirect solar radiation per area (W/m2). The weather station saves all this data on a hard drive every minute for the last five years. 8.1. Instruments: The weather station is divided into two components: a weather mast and a solar tracker. 8.1.1 Weather mast: The weather mast is situated about one hundred meters east of the roof where the other measurements are taken. The weather mast itself measures the air temperature, the relative humidity, the wind speed, wind direction, air pressure, precipitations and solar global radiation. Figure 24: Weather mast located in Jülich 8.1.2 Solar-Tracker: Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 48 Coma can be minimized or eliminated by finding the curvature of the lens surfaces to match the application. /25/ Figure 33: Coma aberration /26/ 9.2 CCD Sensor A charge-coupled device (CCD) is an integrated circuit, which contains a number of linked capacitors. These capacitor arrays are also called photosites, where each represents a pixel in the picture. CCD detectors are based on the photoelectric effect. Incident photons generate electrons by interacting with a semiconductor material. The capacitor array makes up the photoactive region, onto which an image is projected, usually through a lens. The capacitor array is composed of photosites, which accumulate an electric charge proportional to the intensity of the incoming photons. Then a control circuit transfers the charge from one capacitor to the next inside the array. The last capacitor transfers all of its charge into an amplifier, which converts the charge into a voltage signal. This process is repeated until the complete content of the array is converted into a series of voltages. Finally, a digital device samples, digitalizes and stores these voltages inside a memory. CCD sensors can only measure light intensities, in order to obtain polychromatic information, filters have to be used. The use of CCD sensors in astronomy is common because they can be used for a wide part of the electromagnetic spectrum. This makes CCD devices very useful in the work for this thesis, because the solar spectrum is very broad, and it is necessary to capture a large part of it in order to make significant statements about properties of the sun. The camera used in the work for this thesis was a Stingray F-504, which works with a CCD sensor. It is manufactured by the company “Allied Vision Technologies”, and can be seen in Figure 34. In the chart below, the technical characteristics of the camera are given. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 49 Figure 34: Stingray F-504 /27/ Interface IEEE 1394b - 800 Mb/s, 2 ports, daisychain, fiberoptic (GOF) optional Resolution 2452 x 2056 Sensor Sony ICX655 Sensortype CCD Progressive Sensorsize Type 2/3 Cellsize 3.45 µm Lens mount C Max framerate at full resolution 9 fps A/D 14 bit On-board FIFO 64 MB Table 3 : technical specifications of the Stingray F-504 /27/ In addition to the distortions of the telephoto lens, the CCD sensor also has some distortions that can appear in the images. The most typical are the following: Blooming: This effect appears when pictures of very bright objects are captured with a CCD sensor. Each pixel has a limited charge storage capacity. If this limit is reached, the charge flows to its neighbour resulting in misinformation in the final image. This effect is often observed in pictures with very dark edges next to very bright edges. The blooming effect can be minimized with “anti-blooming gates”, which serve as a drainage channels, where the excess charge can flow away. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 50 Smear: In CCD sensors with interlineal charge transfer, the signal of each line of pixels is transferred vertically through a dark area in between these lines. Due to diffraction or scattering, light of a very bright body can extend into the dark transport region. The erroneous information is added to the image during the read-out of the pixel lines, after the exposure is finished. This causes bright lines to appear vertically in the vicinity of intense light sources. A mechanical shutter can prevent this effect, as it prevents light from reaching the sensor during the read-out of the pixels. In Fig. 35 this effect can be seen clearly. Figure 35: picture taken with the camera Stingray F-504 on 29/11/2011 The camera was connected to the computer by Firewire, which can transfer up to 800 Mb/s. The data was managed by the software Allied Vision, the same company that manufactures the camera. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 51 10. Measurements: In the following chapter the measurements taken with the ceilometer, the weather station, the aerosol spectrometer and the CCD camera, in eight different situations, are going to be exposed. The chapter is divided in two different parts: in the first one, all the measurements are going to be listed and commented, and in the second one, the relations between these measurements are going to be studied. 10.1 Results of the measurements 10.1.1. 14th of December 2011 at 13:31 h Figure 36: CHM 15K data on December 14th: rainy day, at noontime the clouds broke up and the sun shot out for a short time In the figure 36 and in the following figures showing the data of the CHM 15K, one can observe three different parameters. The left vertical axis shows the height; in the case of the figure 36 one can observe from 15 meters to 1740. The horizontal axis shows the date and time. The legend situated at the right shows the aerosol particle concentration in the Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 52 air. This data is given by the LIDAR and consists in the quantity of photons per time backscattered by the aerosol particles. /21/ The 14th of December was a rainy day with a very cloudy sky and a varied concentration of aerosol particles and raindrops at a height under 500 m above ground level (see figure 36). The exact time at which the first photo was taken with the CCD camera was at 13:31 h. At this time, the concentration of particles close to ground level, were the lowest of the day. There were some clouds in the sky, but just at 13:31, there was direct solar radiation for a few minutes while at the same time the sunlight was scattered or reflected by the clouds. As there were also some clouds at a height of 3000 m, however thinner than the ones in the figure 36, the weather station registered a high value of diffuse radiation. In the morning there had been some rain, and during the 3 hours before the photo was taken it rained 1.5 mm. The rain gauge recorded 3 mm for the last 24 hours and 13 mm for the last 3 days. This could explain why there was such a low level of particles close to the ground level at the time the picture was taken, because the raindrops pull down the aerosol particles from the air to the ground. In the following table some data from the weather station around 13:31 h is listed. As it can be observed, at 13:30 h the direct radiation was lower, which means that there were clouds in the sky at this time. The low pressure illustrates why the weather was so unstable. Time (hh:mm:ss) RH (%) GR (W/m2) DirR (W/m2) DifR (W/m2) RP (hPa) T (ºC) 13:30:00 70.774 229.534 294.625 123.255 983.03 8.98 13:31:00 70.496 285.444 583.33 127.044 982.98 9.124 13:32:00 70.366 280.953 584.697 126.593 982.98 9.24 Table 4: Weather data on December 14th The quantities of the aerosol particles measured on the ground level were very low, as it can be seen in the figure 37. The exact values at the time of the picture, recorded with the Fidas 200®, are shown in the table below the chart. As it appears in the chart there were some peaks with quantities of almost twenty times more particles with only a few minutes difference. This can happen due to many factors and normally only close to ground level, as a result of the big size of the particles. The representative data should be an average of the data around the time the photo was taken and Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 53 of the measurements of the ceilometer. With this, one can be see how high the concentration of particles in low levels of the troposphere is. Figure 37: Particles concentration on December 14th Time PM 1 PM 2,5 PM 4 PM 10 PM Total 13:28:33 0.6 0.96 1.53 2.46 2.46 13:30:33 0.64 1.02 1.57 1.88 1.88 13:32:34 0.6 0.92 1.23 1.29 1.29 Table 5: Particles concentration on December 14th Finally, the only measurement remaining is the one of the CSR. This measurement is the most important one, but it is also the most difficult to measure. It was not possible to get an exact value of the CSR due to some problems with the devices used. These problems are explained more in detail in chapter 10.2, in the section concerning the comparison of the measurements. Since it cannot be considered as exact, in this chapter, the measurement of the CSR is thus going to be referred as “relative CSR”. In the following chart, one can observe how high the intensity of the sun light in the photo is. It can also be perceived how, outside of the solar disc, the relative intensity decreases very fast, which means that the relative CRS is low. The result of the relative CSR measurement was 22.64%. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 54 Figure 38: Relative intensity of the sun on December 14th 10.1.2. 10th of January 2012 at 14:03 h Figure 39: CHM 15K data on January 10th: cloudy day with a high relative humidity January 10th was a cloudy day but there was no rain. The last time it had rained was the day before with a quantity of 2 mm between 21:00 h and 23:00 h. The last three days there had been 5 mm of rain, and in the previous week about 20 mm. The clouds were situated at a height of 1200 m but the sky was very clear in higher levels of the troposphere. The clouds were not thick enough to reflect all the sunlight, so there was a high amount of diffuse radiation. As it can be gathered from the illustration above, around 14:03 h there was Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 55 not a high level of particles close to ground level. The following table lists some of the measurements from the weather station. The variation among the measurements of the direct radiation indicates the presence of some clouds at this time. Time (hh:mm:ss) RH (%) GR (W/m2) DirR (W/m2) DifR (W/m2) RP (hPa) T (ºC) 14:02:00 73.419 272.294 548.987 104.159 1019.579 9.25 14:03:00 73.659 263.063 527.73 98.94 1019.583 9.304 14:04:00 72.764 263.484 570.815 93 1019.569 9.233 Table 6: Weather data on January 10th The quantities of particles on the ground level measured with the aerosol spectrometer were higher than on the 14th of December, but they were still low. The following diagram and table show the exact data of the concentration of the particles. Figure 40: Particles concentration on January 10th Time PM 1 PM 2,5 PM 4 PM 10 PM Total 14:01:26 2.96 4.49 5.51 9.9 9.9 14:03:26 3.78 5.27 7.77 14.14 14.14 14:05:26 7.59 9.86 12.57 29.86 29.86 Table 7: Particles concentration on January 10th Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 56 Finally, the graph below shows the data of the relative CSR, which seems to be lower than on December 14th, but this will be analyzed more in detail later. The calculated data of the relative CSR was 22.04. Figure 41: Relative intensity of the sun on January 10th 10.1.3. 11th of January 2012 at 11:34 h Figure 42: CHM 15K data on January 11th: clear sky with thin stratus clouds and with a higher concentration of aerosols and a higher relative humidity in the lower atmosphere Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 57 January 11th was a cloudy day, but without rain. The last rainfall had been the night of January 9th, which explains the higher concentration of aerosols at a height between 75 and 300 m in this case. In the morning, at 11:30 h, the sky was really clear, but there were some thin stratus at a height of 7000 m. They are clearly visible in the pictures below. The first one is an image of the ceilometer, which allows observing the atmosphere up to a height of 12480 m. The second illustration is a photo, which was taken with a 3-megapixel camera. It shows how the stratus covered the sun, but also that they weren’t thick enough to reflect all of the direct radiation. Figure 43: CHM15 k on 11th January Figure 44: The sky with stratus clouds on January 11th Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 64 a maximum. The graph is followed by a table, which shows the exact data at the time of the measurements. Figure 51: Particles concentration on January 16th Time PM 1 PM 2,5 PM 10 PM Total 12:13:28 20.17 23.40 31.30 34.44 12:15:29 15.38 18.23 35.98 49.84 12:17:29 13.96 16.82 30.26 37.31 12:43:29 18.67 20.91 30.89 38.54 12:45:29 17.97 20.32 33.59 71.86 12:47:30 18.67 22.10 42.83 90.07 14:43:30 17.52 20.46 30.76 36.15 14:45:29 17.94 21.06 35.61 73.69 14:47:28 18.02 20.52 30.51 43.07 Table 13: Particles concentration on January 16th Finally there are measurements of the value of the relative CSR. On January 16th there were the highest values of relative CSR. As it is seen in the following graphic the values were very similar, at 12:15 h and 12:45 h they were more or less the same value, as it is in graphic, where both lines follow the same way. Relative CSR has been calculated as Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 65 39.31 and 39.58 respectively. At 14:45 h the intensity of the sun was lower, but in the graphic the curve is similar than the others, this happened because the weather and air conditions were similar. The value of relative CSR here was 35.31. Figure 52: Relative intensity of the sun on January 16th 10.2. Relation between the measurements: In the following pages, the different measurements which were taken are going to be compared and analysed. All the collected data will be included in this analysis in order to identify which measurement has an impact on the relative CSR. As has already been mentioned in the beginning of the chapter, the CSR is referred to as relative CSR because the devices used for the measurements of this thesis could not provide the accurate data necessary for calculating the exact CSR. The first problem, which makes the measurements inexact, is the light in the sky that is not part of the direct radiation. The CCD sensor captures light from the dark sky behind the sun and from the diffuse radiation, so this light should not be included in the circumsolar radiation or in the sun radiation. If this radiation is not considered the calculated CSR will be higher than the real one, that is because the intensity inside the solar disc is not influenced by this radiation (is already the maximum that the sensor can measure), but it is added to the circumsolar area’s light. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 66 The size of the sun in the pictures is another problem found in the measurements. The entire sun diameter was inside the picture; its diameter measures 1227.26 pixels and the length measured by the program is 1748 pixels. As one can observe in the figures where is shown the relative intensity of the sun, the intensity is still decreasing at the end of the measurements, which means that there is light from the circumsolar radiation that is not included in the calculated CSR. That makes an estimated CSR lower than the real one. As has been written in previous chapters, the intensity of the sun should be higher in the area around the centre of the sun than at the outer edge of the solar disc. However, the measurements taken for the research of this thesis list a constant data for the entire solar disc, which means that the calculated data of the CSR might be imprecise. But by observing the changes in the circumsolar area one should be able to come to significant conclusions all the same. The direct radiation measured by the pyrheliometer refers to the direct radiation plus the radiation of the circumsolar area, which is why the total intensity recorded with the CCD sensor, should be related to this data. The following table lists the measurements of the direct radiation and of the diameter of the circle with maximum intensity, taken at the same time. Day Diameter (pixels) Direct Radiation (Wm-2) December 14th 649 584 January 10th 631 550 January 11th 613 474 January 13th (11:06) 631 554 January 13th (12:17 h) 637 560 January 16th (12:15 h) 691 730 January 16th (12:45 h) 697 761 January 16th (14:45 h) 640 628 Table 13: Diameters and direct radiation As one can observe, the diameter of the circle with maximum intensity is normally longer if there is more direct radiation. This influences the relative CSR data because the intensity inside the solar disc should be higher than in the circumsolar area. This makes the relative CSR calculated in the days with high radiation to be higher. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 67 In order to compare the data, it is necessary to keep in mind that there are other factors that could change the intensity of the sun. They will be commented on later. As can be seen, the data follows a logical order, linking the intensities and the diameters. However, there is one exception, which becomes obvious in the comparison of the 14th of December and the 16th of January. These days were really different; they didn’t have any measurement in common; temperature, relative humidity, atmospheric pressure, diffuse radiation, level of particles and relative CSR were different. The following pages will give further details of the comparison of all these factors. Regarding the temperature, it is similar for most of the days: 2.7ºC being the coldest and 9.7ºC the warmest result. The temperature is important because the maximum level of the absolute humidity depends on it. The atmospheric pressure affects the weather, but not directly in terms of the amount of direct or indirect radiation. The 14th of December was the only day there was rainy and unstable weather and also low pressure. On the other days, there were high pressures and stable weather. The relative humidity is more important in relation to its effect on the solar radiation. A higher humidity in the air means that there is also more water vapour, so this water can diffuse the solar radiation. With a higher level of humidity in the air the diffuse radiation should be higher than with a low humidity under the same conditions, which means the same should apply to the relative CSR. In the table below shows the data of the relative CSR, the indirect radiation and the relative humidity. Day Intensity diffuse radiation (Wm-2) Relative humidity (%) CSR 14th December 127 70.5 22.64 10th January 100 73 22.04 11th January 42 80 24 13th January (11:06 h) 56 84 32.9 13th January (12:17 h) 103 77 33.64 16th January (12:15 h) 7.7 50.6 39.31 Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 68 16th January (12:45 h) 10.8 71 39.58 16th January (14:45 h) 0 58 35.31 Table 14: data of the relative CSR, the indirect radiation and the relative humidity The obtained results listed above make it difficult to identify the impact of the relative humidity. But on the 16th of January, there were similar conditions throughout the whole day and, most importantly, there were no clouds in the sky. Between 12:15 h and 12:45 h there was a 20% increase of the relative humidity and it as one can see in the table, the CSR and the diffuse radiation were a bit higher. Concerning the diffuse radiation and the global radiation, the amount of these radiations will be analyzed, as well as why they exist and if there is any relation between the diffuse and global radiation and the direct radiation. As has already been said in previous chapters, the sunlight is scattered by small particles in the air. Normally, the majority of the diffuse radiation comes from the direct radiation scattered by the clouds and other aerosols particles. The majority of the clouds that scatter the solar radiation are situated between 800 m and 2000 m. The scattered radiation doesn’t only come from the clouds that cover the sun from the point of view of the observer. It also comes from the other clouds that diffuse the light in all directions. When low clouds are covering the sun, part of this diffuse radiation can be registered in the pyrheliometer as direct radiation, because the scattering can be very light, and be part of the circumsolar area. On days with clear skies, there is also some diffuse radiation because the aerosol particles scatter the light as well. The following charts detail the diffuse radiation on the days the measurements were taken, as well as the comparison between diffuse, global and direct radiation on the 11th of January. In the first chart, the measurements of diffuse radiations of every day are shown. They were not very regular on cloudy days, and when the sky was completely overcast, the diffuse radiation did not exist or was very low due to the high albedo of the clouds. In the second chart, one can see how the different radiations changed in an aligned manner on the 11th of January. When there was no direct radiation, the diffuse was lower, especially when the sky was completely overcast. The global radiation was more constant, but always increased or decreased depending on the amount of direct radiation. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 69 Figure 53: Diffuse radiation each day Figure 53: Global, diffuse and direct radiation on 11th of January Now the relation between aerosol particles and radiation is going to be examined. In order to carry out the research correctly, the weight of the small particles measured has to be separated in coarse and fine particles. As has already been mentioned previously, the particles with the greatest impact on the solar radiation are the small particles, with a size under 2.5 PM. The chart below shows the quantities of particles with a size bigger than 2.5 PM in relation to the global, diffuse and direct radiation on the 16th of January. One can observe that there is no relation between coarse particles and radiation. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 70 Figure 54: Coarse particles and radiation on January 16th At 12:15 h, the amount of particles bigger than 2.5 PM was between 10 and 30 g/m3, and at 12:45 h between 17 and 67 g/m3. But the calculated data of the relative CSR was very similar: 39.31 and 39.58. While the particle data shows great differences, the concentration being up to 10 times higher at times, this variation does not affect the radiation level. Finally, the following charts illustrate the comparison of the different relative CSR measurements. The different graphs in the diagram below show the variation of the intensity outside of the solar disc in all of the measurements that were taken. Figure 55: Relative intensity in the circumsolar area each day Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 71 At first sight, two different groups can be distinguished: the measurements on days with low particle levels (December 14th, January 10th and January 11th), and the rest, with high levels of particles. The most important difference is the way in which the level of the sun intensity in the circumsolar area decreases. As had been mentioned earlier, the days with more direct radiation had a bigger intensity around the solar disc, but this didn’t mean that the relative CSR was bigger. It is thus going to be more representative how the intensity decreases outside of the solar disc. In the following, the measurements will be taken in pairs and then put in comparison in order to determine these differences. The first comparison will be between the measurements of the 14th of December and the 13th of January at 12:17 h. They have a similar data of direct radiation and also a similar diameter with top intensity, so it is going to be easy to compare them. On the 14th of December the average weight measured of particles under 2.5 PM on the ground level was around 1 g/m3 and on the 13th of January it was 8,8 g/m3. The ceilometer measured a concentration of 5-20 below 500 m on December 14th and of 100-220 on January 13th. There was thus a considerable difference between the results. The relative CSR was also very different. In the following table, the variation of the relative intensity can be observed: Distance (pixels) / Day December 14th January 13th (12:17 h) 18 100 % 99.99 % 36 99.59 % 95.97 % 42 85.69 % 95.47 % 48 64.84 % 92.02% 60 42,2 % 83.14 % 72 31.08 % 72.72 % 102 18.66 % 44.72 % 150 14.42 % 26.23 % 228 9.86 % 9.84 % Table 16: comparison of solar relative intensity in the circumsolar area at December 14th and January 13th The distance is measured from the border of the solar disc. Around the sun the intensity was relatively the same. One could remark slightly higher numbers on the 14th of December, which decrease very quickly with greater distance to the sun, which indicates a small Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 72 CSR. The intensity on the 13th of January, on the other hand, decreases much more slowly, which means that there is a higher CSR. On the 13th of January at 12:17 h and on the 16th of January at 14:45 h the diameters with maximum intensity were of a similar size. The concentration of small particles on the ground level was very different; on January 13th it was 8,8 g/m3 and on January 16th it was more than twice as much: 20 g/m3. Regarding the ceilometer, the concentration on January 13th was between 100 and 220 at a height below 500 m and towards the ground level it became increasingly higher. On the 16th of January the concentration was similar, but the highest levels were above 75 m. The following table shows the relative intensity measured from the border of the solar disc. Distance (pixels) / Day 13th of January (12:17 h) 16th of January (14:45 h) 32 99.85 % 99.99 % 42 95.37 % 98.37 % 60 85.65 % 90.62 % 90 69.71 % 74.56 % 120 55.55 % 59.33 % 200 20.83 % 25.18 % Table 17: comparison of solar relative intensity in the circumsolar area between January 16th and January 13th In this table one can observe that the graphs from the two days in question are quite similar, and that their relative CSR data doesn’t vary much from another. The main differences one could highlight are in the solar intensity and in the PM on the ground level. The next comparison will be between the 10th and the 11th of January. These days were much more similar than the previous ones in relation to the particle levels and the weather conditions. On the 10th of January, the concentration of particles with a size under 2.5 PM on the ground level was measured about 5.5 g/m3, very similar to that of the 11th of January, which was about 5m/m3. However, the particle concentration measured by the ceilometer showed some differences. At low heights (below 200 m), the concentration measured on January 10th was between 5 and 40, and on January 11th between 20 and 100. The amount of particles was thus similar at the ground level, but not at a higher levels. One should also not forget to consider the influence of the cirrus clouds at a height around 7000 m on January 11th. The other big difference was in the direct solar radiation, which is Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 73 why the length of the diameter with maximum intensity was also different and it was harder to see the variation of the relative CSR between both days. The best way to perceive the variation is by observing the different distances (pixels) in relation to the decreasing intensity. The table below illustrates this and it is clearly visible that the relative CSR is higher on 11th January. Relative intensity / Day January 10th January 11th 100 % to 95% 13 pixels 17 pixels 95% to 90% 4 pixels 6 pixels 90% to 80% 6 pixels 12 pixels 80% to 65% 9 pixels 18 pixels 65% to 50% 9 pixels 18 pixels 50% to 25% 27 pixels 42 pixels 25% to 18% 24 pixels 24 pixels 18% to 10% 120 pixels 120 pixels Table 18: comparison of the relative solar intensity in the circumsolar area between January 10th and January 11th One has to remark that the diameter with maximum intensity on the 10th of January was 18 pixels larger than that on the 11th of January and that this table thus only considers the variation of the light intensity as the measurements were not taken in the exact same place. Annex A1 Calculations The appendix includes the measurements that were taken in order to calculate the estimated value of the circumsolar ratio. As has been said in previous chapters, the equation used to calculate the direct beam of the sunlight and the circumsolar radiation depends on the size of the sun and the light intensity. It was thus first of all necessary to know the exact size of the sun in the photos. To make this possible it was required to calibrate the camera. In order to calibrate the camera, pictures of another object had to be taken. The same focus had to be used for this object as for the sun, so it needed to be located far away. From the roof of the SIJ (the place where the photos of the sun were taken), the tower of a church was visible on a clear day. The distance between the SIJ and the church was calculated by coordinates and it was 3374.11 meters. Then, it was necessary to know the size of the church in order to compare it to the size of the sun. The church, which had a prism shape, was measured at the site and its side length was 5.58 meters. With the data of the sun diameter and its distance to the earth, the size of the sun in the photos could be calculated. The software Visual Builder was used to measure the size of the church in pixels. The programme recognises the points where there are any changes in the photos, then the lengths between these points can be measured in pixels. The figure 57 shows the points detected by Visual Builder. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 82 Figure 57: Measurement of the left side of the church. For the comparison of the size in pixels and metres, the measurements of both sides of the prism were necessary in order to know the angle from which the photos were taken. The figure 58 explains it. Figure 58: Sides of the church and angles Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 83 With the measurements from both sides (left and right), the angle can be calculated with trigonometry. Then, one can calculate the data of the visible side in meters. To obtain this data, the following relations were used: sinφ = a left / a (1) cosφ = a right / a (2) Dividing (1)/(2) φ = arctg a left / a right a left (metres) = sen φ * a Equation 12: trigonometry relations The data 'a' right and left can be calculated knowing the data of the angle. Then it only remains to compare the size of the church in pixels (Pc) with the real size of the church in metres (Sc) and the distance between the camera and the church in metres (Dc) as well. The size of the sun radius (Ss) and the distance between the sun and the earth are also known (Ds), so the size of the sun in pixels (Pd) can be calculated with the following relation: Equation 13: length relations The data of the sun radius is about 6.96 E9 m, and its distance to the Earth in December - January (the orbit is elliptical) is about 1.4710556 E11 m. This distance is called periapsis and it is the closest the earth and the sun can be. According to the estimations of scientists, the periapsis is on the 4th of January. The measurements for this thesis were taken between December 16th and January 13th, and the distance to the sun in these days should be very similar to the periapsis, therefore this length was used. In the table 19, one can observe the data and the final results of the measurements taken in order to calculate the sun radius in the photos. Three different measurements were taken, so that a better result would be obtained. Constant data Sun radius 6.96 E9 m Distance earth-sun 1.4710556 E11 m Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 84 Distance camera-church 3374.11 m Church side 5.58 m First measurement Right side measured 42.92 pixels Left side measured 214.98 pixels Angle obtained 78.7 º Left side obtained 5.859 meters Sun radius obtained 614.04 pixels Second measurement Right side measured 42.02 pixels Left side measured 214.62 pixels Angle obtained 78.92 º Left side obtained 5.593 meters Sun radius obtained 612.63 pixels Third measurement Right side measured 41.58 pixels Left side measured 215.33 pixels Angle obtained 79.07 º Left side obtained 5.5966 meters Sun radius obtained 614.21 pixels Sun radius average 613.63 pixels Table 19: results of the measurements Once the sun size has been calculated, the sun radiation in the whole circumsolar area needs to be measured. With the devices used, it was not possible to calculate the exact incoming solar radiation from the different points. It was, however, possible to distinguish it due to the different intensity of the light in the circumsolar area. The variation of the solar radiation is thus determined by its brightness. The following equations were used to calculate the CSR: Equation 14: CSR equation Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 85 Equation 15: intensity of the circumsolar area or the solar circle. The difference between these equations and the equation 15 in the third chapter of the thesis is that the flux is considered as brightness per area instead of sun radiation. In the pictures taken, one can observe the decreasing intensity at the border of the sun. This variation could be measured with the software Visual Builder. In the figure 59, one can see one of the photos used for the measurements in this thesis. Figure 59: Picture of the sun on January 13th A programme, which was created for the purpose of previous studies by the SIJ, can measure the light intensity in a picture for different areas. The programme then exports the information to an excel file. As one can observe in the figure XXX, the largest possible diameter one can obtain in the picture (due to the focus of the tele lens), is the diagonal one. The diagonal diameter has thus been used to measure the variation of the light intensity. The following picture shows the different areas that the programme used to measure the light intensity. It measured the average intensity for each area inside of the green rectangles. One can see that the rectangles are smaller at the edges of the sun circle because it is the most important area. Analysis of aerosol particles for the determination of energy losses in solar power plants Juan Prats Briceño 86 Figure 60: Picture obtained from the software Visual Builder, with which the sun on January 13th 2012 is measured. The excel file positions every rectangle in the picture, providing all of the distances in pixels. In order to calculate the estimated circumsolar radiation, it is required to know the brightness of the solar circle and the brightness included into the circumsolar area. Therefore, for the determination of the brightness inside the sun circle, it is necessary to add the flux (area*brightness) inside a circle with a radius of 613,63 pixels. It is further necessary to also add the flux of the circumsolar area. The CSR can then be calculated. The table 20 shows part of the results of the measurement on January 13th. As one can see, it shows the average intensity and the centre of each rectangle. Intensity R95 Center.X Position (Pixel) 1699 Intensity MR95 Average Intensity 107,72916 Intensity R96 Center.X Position (Pixel) 173 Intensity MR96 Average Intensity 101,63125 Intensity R97 Center.X Position (Pixel) 1705 Intensity MR97 Average Intensity 101,38541 Intensity R98 Center.X Position (Pixel) 167 Intensity MR98 Average Intensity 95,08334 Intensity R99 Center.X Position (Pixel) 1711 Intensity MR99 Average Intensity 95,25625 Table 20: data obtained from the Visual Builder All of the calculations necessary to obtain the CSR data, which is specified in the chapter 10 “Measurements”, were done with the software Microsoft Office Excel.