留學生assignment：Performance of a 500 kWP grid connected photovoltaic system at Mae Hong Son Province, Thailand

Contents
Abstract I
Keywords I
I Introduction 1
II System component, design and feature 2
III Monitoring system 2
IV Evaluation Of Community-Based Activities 3
V Solar high-temperature high-flux treatment of hazardous wastes 3
VI Materials and methods 4
VII Conclusions 7
References 7
Acknowledgements 8
 
Abstract: This paper summarizes the first eight months of monitoring of the PHA BONG photovoltaic Generation project, a 500 kWp photovoltaic pilot plant, in Mae Hong Son province, Thailand.The local grid in this remote area in the North West of Thailand is very limited in its capacity and cannot be enlarged. It has been in operation since 20 March 2004 by feeding into 400 VAC, 22 kV medium voltage grid. The system consist of a photovoltaic array 1680 modules (140 strings,12 modules/string; 300 W/module), power conditioning units and battery converter system. During the first eight months of this system’s operation, the PV system generated about 383,274 kWh. The average of generating electricity production per day was 1695.9 kWh. It ranged from 1452.3 to 2042.3 kWh. The efficiency of the PV array system ranged from 9 to 12%. The efficiency of the power conditioning units (PCU) is in the range from 92 to 98%. The final yield (YF) ranged from 2.91to 3.98 h/d and the performance ratio (PR) range from 0.7 to 0.9.

Keywords:  Solar power system; Performance; Grid connected system

I Introduction
Mae Hong Son province is situated in North Western of Thailand on the bounder with Myanmar. The Province has only 22 kV distribution lines which take power from Chiang Mai substation and passes through trees and hills for 250 km. The supply often fails, as trees touch the conductors. The main electricity supply of Mae Hong Son, in the province’s amphora mange zone, has three sources, The Pha Bong dam (PB-Dam;1850 kW), The Mae Sa Yha Dam (MSY-Dam; 2!3375 kW), and Mae Hong Son Diesel generator (MHS-Diesel; 31000 kW,3!1250 kW). The total generation capacity of this installation is about 14,350 kW. Normally power generation from dams is able to supply electricity about eight months per year, and is shut down in summer and winter seasons. The PB-Dam can supply about 460 kW and the MSY-Dam can supply about2000–4000 kW. Generation capacity depends on the season as the dams were not built for storage. The MHS-Diesel generator can only supply about 6000 kW. The policy of the Electricity Generating Authority of Thailand (EGAT) gives first priority to generation from the dams, a renewable energy resource. If the supply from dams is insufficient, diesel generation used to supplement the supply. In this way the use of costly diesel oil is minimized. The peak demand on 10 October 2004 at 19.30 p.m. was about 4660 kW and the maximum day-load at 15.00 p.m. was about 3030 kW. The supply constrains#p#分頁標題#e#
require careful planning of the electricity supply in Mae Hong Son.EGAT, the organization responsible for supply and generation of electricity in Thailand, encourages the study, exploration and planning for use of renewable energy in this province. It has initiated a pilot PV project in Mae Hong Son, the 500 kW Pha Bong photovoltaic generation. The project has three objectives; to increase power supply, to decrease consumption of diesel fuel during daylight hours and finally to encourage a national strategy in the production photovoltaic cell and accessories. Work on the Pha Bong photovoltaic generation was commenced on 11 February 2003. The system first began to supply the first electricity to 22 kV grid system of Provincial Electricity authority (PEA) on 20 March 2004 and has been completely operational since 24 March 2004. The project was handed over to EGAT on the 9 April 2004. However, far the performance of their paper data was collected from 24 March 2004 when the project was completely operational and being fully monitored until 31 October 2004 [1–3]. This paper presents an evaluation of the performance of system during the first eight months of operation. The performance of the components of the system (PV arrays, power condition unit) is analyzed, and finally the performance of the whole system is investigated. A later paper will review information about the performance of the battery and converter system used to improve the reliable of the system and its connection to the grid.

II System component, design and feature
The PHA BONG photovoltaic generator has a total power capacity of 500 kWp. It consists of a 1680 PV modules (140 strings, 12 modules/string, 300 W/module). The generator is divided into two, 250 kWp, sub-arrays of double glazed ASE-300-DG-FTmodules from RWE SCHOTT Solar. There are 1.28 m wide, 1.90 m long and a face south and are tilted at 158. For grid coupling two power conditioning unit (PCU1, PCU2) each with a nominal power of 250 kVA are used. The Inverters function according to the new sunny team principle ensuring a high reliability due to the optimized efficiency in the lower part-load range. Two bi-directional battery inverters (BC1 and BC2), each with a power output of 200 kVA, are operated in parallel. The battery inverters are connected to a battery bank (280 pcs 2 V/pcs; 560 V, total 1200Ah) and can feed into the grid in addition to the PV power. A drastic and rapid change of the grid feeding power, for example if the PV array is shaded by cloud, is avoided by using fast microprocessor—based compensation of the battery inverter. Batteries are charged between 22.00 p.m. and 06.00 a.m. If the PV array cannot produce, the batteries will discharge continuously to system in a short time less than 5 min. The PCUs and battery converters have their own operational control and can be operated independently of the system controller’s status. To realize the grid connection an AC junction box (AJP) is used. For visualization, data logging and some operations a system controller and two operating personal computers are used.[4].#p#分頁標題#e#

III Monitoring system
The objective of solar chemical engineering and materials research is to establish useful applications of concentrated solar radiation for processes in the chemical, oil and gas converting, smelting and ceramic industries. Eventually, feasible procedures should be developed to store solar energy as a “solar chemical fuel”, in order to make solar energy available regardless of time and location. Fossil fuels, which are required for the conversion of crude feedstock’s or for the production of basic chemicals, could be substituted by solar energy as well if the costs are acceptable. For the short- to mid-term, future market niches include the solar photochemical production of specialty chemicals and solar detoxification of polluted water and hazardous wastes. Concentrating solar technologies can also be used beneficially for the manufacture, testing and heat treatment of materials. This paper is a short review of research and development (R&D) areas in solar chemical engineering as they are exploited within task 3, “Solar Chemistry and Solar Materials Research”, of the “Solar Energy Association North Rhine–Westphalia, Germany”. Results of the first five-year period of research in solar chemistry and materials have been presented in a recent volume [5]. Thus, this paper the highlights progress that has been achieved in some of the projects since then. Because the direct transfer of laboratory results to pilot applications is a risky enterprise, a high-flux solar furnace was constructed in 1993/4. It started operation in summer 1994 at the DLR research center in Köln-Porz, Germany. The furnace is supposed to bridge the gap between laboratory experiments and large-scale pilot or demonstration experiments. If the furnace experiments are successful subsequent scale-up can be carried out at solar test centers, such as the Platforms Solar de Almer??a in Southern Spain. In these centers large solar concentrating systems are available in the 100 kW to multi-Megawatt range. For both industrial clients and academic researchers, the furnace serves as the tool to develop and to qualify new techniques in solar chemistry and materials research at relatively low expense

IV Evaluation Of Community-Based Activities
There is a chance that solar energy could contribute to environmentally benign and sustainable development in the process industries of the 21st century. It is necessary to collect further experience in the operation of solar chemical processes and to test those new techniques on a small engineering scale that are already running in the laboratory[6]. The DLR solar furnace has been placed at disposal for the required tests. Technology developments include receiver/reactors, windows for the receiver/reactors, final concentrators, and specific high-temperature high-flux measurement techniques. The main features of volumetric and direct absorption receiver/reactors are that energy transfer is by mass less high-flux radiation and that feasible heat flux densities are more than one order of magnitude higher than in conventional reactors. Direct use of solar radiation instead of artificial lamps is highly attractive for the photochemical production of industrially important chemicals. Photochemical syntheses could easily be developed into early applications of solar radiation in chemical technology[7]. The opportunities for the industries of the industrialized world are twofold: solar chemical techniques could be employed directly at locations in industrialized countries if there is enough sunshine available or could be exported to those industrializing countries that have good solar conditions. In a future world economy, these countries could export chemically stored solar energy instead of oil and natural gas or they could export solar processed basic chemicals instead of the raw mineral material[8].#p#分頁標題#e#

V Solar high-temperature high-flux treatment of hazardous wastes
A main objective of waste treatment is to transfer the hazardous ingredients to an environmentally benign form to allow safe disposal. Toxic organics must be destroyed and leachable in organics must be fixed. A second objective is to transfer the valuable components to secondary raw materials in sufficient quality for reuse[9]. High-temperature treatment has a very high specific energy demand which is conventionally met by combustion of fossil fuels. Because the gases of combustion are polluted by volatiles of the waste materials, related processes have to purify big exhaust gas streams. The investment cost for exhaust gas treatment may be higher than 50% of the total investment. Use of concentrated solar radiation for thermal waste treatment enables substitution of fossil fuels. Lower off-gas streams occur and the generation of carbon dioxide, caused by combustion of the fuels, is avoided. Thus, lower investment and operation costs are expected for off-gas purification. In a solar waste treatment, recycling or remolding plant wastes are irradiated directly with highly concentrated solar radiation in a directly absorbing receiver/reactor that integrates two functions: the conversion of high-flux radiation to heat and physico-chemical conversion of the material.
To examine the feasibility of solar thermal waste treatment and remolding of aluminum scrap, a direct absorbing rotary kiln receiver/reactor was constructed on a mini-plant scale for tests in the solar furnace. Design details, construction features of the mini plant and draft flow sheets for pilot- and technical-scale configurations have been published elsewhere and. Solar thermal recycling of aluminum scrap was selected as a reference case to prove the feasibility of the receiver/reactor concept, because extended use of aluminum will lead to a further growth in aluminum production and to an over-proportional growth in secondary aluminum. Temperature levels above 1000°C have been reached in the rotary kiln[10]. and model salt mixtures and aluminum were molten. Successful melting experiments make us optimistic regarding the feasibility of the process. It is the purpose of the running project to show both feasibility and to estimate costs that might be expected after scale-up. Another solar melt-pot receiver/reactor for beam-down solar irradiation was developed for a variety of melting processes up to 1300 K, amongst them aluminum recycling.

VI Materials and methods
The organic dyes used in this research are three different derivatives of the parent molecule, 4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (BODIPYs), obtained from our collaborative partner, Molecular Probes Inc. They areodesy 494/505, BODIPY 535/558, and BODIPY 564/591 to be referred to in this paper as dyes A, B, and C,respectively, to remain consistent with previous work. Their absorption coefficients are 65600, 76400, and 78000 cm_1M_1, respectively. This and other pertinent physical and spectral properties of these dyes have been previously reported . These dyes are characterized by strong absorption peaks in the visible region, high fluorescence quantum yields, weak interaction among like molecules, and moderate-to-small overlaps of their individual absorption and fluorescence spectra making them excellent candidates for use in LSCs. LSCs can be made from a wide variety of materials. For this study, we focused our attention on LSCs made of thin films of a methyl acrylate/ethyl methacrylate co-polymer (n   1.49) that are solvent cast onto clear glass substrates (n   1.52). Thin films have the advantages of being simple to fabricate and able to accommodate the high dye concentrations needed for efficient FRET to occur. To make dye-doped polymer films, dyes were obtained incrystalline form, weighed, and then dissolved in toluene to make stock solutions that were stored in the dark until further use. Similarly, stock solutions of polymer in toluene were also made. After all solutions were thoroughly mixed, the concentrations of dye in each of the stock dye solutions were calculated using the Beer–Lambert Law. Solution absorption spectra were measured from 350 to 650nm with a bandwidth resolution of 1 nm using a UV/visible absorption spectrometer (Cary Win-UV/VIS BIO 300 Spectrophotometer). Dye–polymer solutions were made by adding a specific volume of a stock dye solution to a specific volume of the stock polymer solution to yield the target dye concentration of 1_10_2M in the polymer film. Six solutions were mixed in all, one for each individual dye and three for each combination of the different dyes. Each dye–polymer solution was stirred in a vortex mixer for 20 min and stored in the dark for 24 h before casting. Thin films were statically cast by pipetting a specified volume of a dye–polymer solution onto a glass substrate. After drying, the optical density (OD) of each film was measured. The thin films were designed to maintain the same OD for each dye in every plate. For example, the LSC plate that contained dyes A, B, and C (ABC-LSC) had the same OD for dye A as the LSC plate that contained only dye A (A-LSC). The OD’s for the dyes were chosen to achieve high absorption of light at their peak values. They were reasonable and experimentally convenient choices for LSCs, but may not be the optimum ODs for maximum LSC performance. All absorption spectra were measured directly from films deposited on 10 cm_10 cm_0.2 cm glass substrates. For fluorescence measurements, the LSC plates could not be used because of reabsorption artifacts [11]. To prevent these artifacts, fluorescence and fluorescence excitation spectra were measured from thin films with ODs less than 0.05 that were cast on glass slides of dimensions 3 cm_1 cm_0.05 cm.These films were cast from the same dye–polymer solutions used for the 10 cm_10 cm_0.2 cm plates. The thicknesses of the films were calculated via the Beer–Lambert Law to be approximately 15 mm for the 10 cm_10 cm_0.2 cm LSC plates and 0.5 mm for films on glass slides used to measurefluorescence spectra[12].. Fully corrected fluorescence and fluorescence excitation spectra were measured with a research-grade fluorimeter (Photon Technology International, Model QM-2) as previously described . Fluorescence was measured from the front surface of the slides by orienting their normal to 301 from the excitation beam . Appropriate long-pass filters were placed in the emission beam path to block any exciting light from being detected. Collection bandwidths for fluorescence spectra were 4 nm for excitation and 2 nm for emission with the numbers reversed for measuring fluorescence excitation spectra. The anisotropy of fluorescence was measured as previously described [11]. The absorption and fluorescence spectra of multiple-dye films were deconvoluted using a sum of least-squares fitting routine to obtain each dye’s contribution to the spectra. To measure the light output from a 10 cm_10 cm_0.2 cm LSC, one edge of the plate was placed against two PVCs (Solar World, SuperCell) that were positioned next to each other to form a detection area of 12-cm long by 2-cm high. The two PVCs were wired in parallel and connected to a digital voltmeter to detect open circuit current. The other three sides of each LSC were blackened with electrical tape to eliminate artifacts caused by reflections off of these surfaces. A light-tight box containing the PVCs shielded the cells from any excitation light that might have fallen directly on them to less than 0.001mA generated. Current measurements from the PVC detector reflect relative changes in the light outputs at the edges of the LSCs and should not be confused with measurements of maximum power performance of the LSC–PVC system. The coupling of the PVCs to the LSC plates was not optimized, but was done in a repeatable manner to ensure good reproducibility. Any variations in intensity of emission across the edge of the plate and in the angle of refracted emission are the same for all the LSCs examined within the accuracy of our measurement. A solar simulator (Spectral Energy Corp.) was used as the excitation source for measuring the outputs from the LSCs. It emitted an AM 1.5 global solar spectrum with a spatially and temporally averaged intensity of 68 mW/cm2. A titanium oxide white background of the same area was placed 2mm below a LSC plate inserted into the PVC detector system. This allowed the incident light to make a second pass through the LSC, making the effective OD approximately double that of the plate’s measured absorption spectrum . The percentage of incident light absorbed by a LSC was calculated using its absorption spectrum, its effective OD, and the AM 1.5 global solar spectrum [13]. The percent efficiency of FRET from one dye to another was determined by taking the number of excitations created in the dye being transferred to (the acceptor) and dividing it by the number of photons absorbed by the dye doing the transferring (the donor) . The number of excitations created in the acceptor dye was measured by the fluorescence excitation spectrum of emission from the acceptor dye. The percent efficiency was calculated by normalizing the fluorescence excitation and absorption spectra at a wavelength where only the acceptor had appreciable absorption and then dividing the former by the latter. By definition, a photon absorbed directly by the acceptor was equivalent to 100% energy transfer, so this is why the two curves were normalized here . A second measure of FRET was determined through the deconvolution of the fluorescence spectrum into components from each dye when using exciting light absorbed primarily bythe donor.#p#分頁標題#e#

VII Conclusions
The merit of the strategy of increasing the absorption of solar light by a LSC through the use of many dyes is verified by a marked increase in output of multiple-dye LSCs over single-dye LSCs. Highly efficient FRET between multiple dyes is shown to optimize the output of a multiple-dye LSC. Maximizing the fluorescence quantum yield and minimizing the reabsorption losses of the final dye in the network is of critical importance in defining the optimal performance of the FRET LSC. The challenge in making an efficient multiple-dye LSC is to find dyes that have suitable properties to make such a system work. They must be soluble at high concentrations in the same solvent used to cast the polymer and they must form a good FRET network.

References
[1]  M. Becker, K.-H. Funken, Editors , Solare Chemie und Solare Materialforschung, C.F. Müller Verlag, Heidelberg (1997).
[2] M. Becker, M. Böhmer, K.-H. Funken and A. Kalt, Der Hochflußdichte-Sonnenofen — ine neue Testanlage der DLR, M. Becker, Editor, Der Hochflußdichte-Sonnenofen, DLR, Köln (1994), pp. 17–22.
[3]  M. Becker, M. Böhmer, Editors , GAST The Gas-Cooled  Solar Tower Technology Program, Springer-Verlag, Berlin (1989).
[4] A. Neumann and A. Witzke, The influence of sunshape on the DLR solar furnace beam. Sol. Energy,  66  (1999), pp. 447–457.
[5] K.-H. Funken, G. Luedtke, J. Ortner and K.-J. Riffelmann, Solar photochemical production of ε-caprolactam using highly concentrated sunlight. J. Inf. Recording,  25  (2000), pp. 3–14.
[6] K.-H. Funken, F.-J. Müller, J. Ortner, K.-J. Riffelmann and C. Sattler, Solar collectors versus lamps — a comparison of the energy demand of industrial photochemical processes as exemplified by the production of ε-caprolactam. Energy,  24  (1999), pp. 681–687.
[7] C. Sattler, F.-J. Müller, K.-J. Riffelmann, J. Ortner and K.-H. Funken, Economic evaluation and comparison of the industrial photosynthesis of ε-caprolactam via solar or lamp operated photooximation of cyclohexane, R. Hogan, Y. Kim, S. Kleis, D. O'Neil, T. ASME/KSME/JSME/SAREK/JSES/KSES Conference on Renewable and Advanced Energy Systems for the 21st Century, Maui, USA, 11–15 April 1999, The American Society of Mechanical Engineers ASME, New York (1999) (paper no RAES99-7692).
[8] K.-H. Funken, B. Pohlmann, E. Lüpfert and R. Dominik, Application of concentrated solar radiation to the high temperature detoxification and recycling processes of hazardous wastes. Sol. Energy,  65  (1999), pp. 25–31.
[9] B. Pohlmann, K.-H. Funken and R. Dominik, A solar heated rotary kiln for the detoxification of hazardous wastes. J. Phys. IV France,  9  (1999), pp. 307–312.
[10] H. Schmidt-Traub and T. Hahm, Application of computational fluid dynamics to solar thermal receiver/reactor modeling. J. Phys. IV France,  9  (1999), pp. 147–152. Advancing health disparities research: can we afford to ignore measurement issues? Med Care. 2003;41(11):1207–1220.#p#分頁標題#e#
[11] J. Lakowicz, Principles of Fluorescence Spectroscopy, Plenum Press,
New York, 1983.
[12] B.P. Wittmershaus,  J.J. Skibicki, J.B. McLafferty, Y.-Z. Zhang,
S. Swan, J. Fluorescence 11 (2001) 119.
[13] American Society for Testing and Materials, Solar Spectral Irradiance Standard Curves, international standard curve for Terrestrial Global Air Mass 1.5 ASTM G173-03, 2003,
Acknowledgements
I thank my supervisor, Dong Liangjie for she excellent help in the process of my thesis writing and Mr. Dang, for giving English Knowledge and further information.