An Innovative Gas Turbine Plant for Parabolic Trough Concentrated Solar Power Plants

Transcription

1 An Innovative Gas Turbine Plant for Parabolic Trough Concentrated Solar Power Plants Roberto Cipollone, Andrea Cinocca. Department of Industrial and Information Engineering and Economics, University of L Aquila, L Aquila, 67100, Italy E -mail: roberto.cipollone@univaq.it, andrea.cinocca@univaq.it Abstract - Concentrated Solar Power Plants (CSP) technology using Parabolic Trough (PT) has the capability to give, in the mean future, a strong contribution to the electrical energy generation. In the long term, inside a new framework of relationships concerning energy production, many aspects would justify a significant contribution to the phase out of fossil sources use. The paper concerns about a theoretical modeling aimed at improving the performances of CSP which approaches the energy generation from a system point of view. Thanks to it, the attention is focused on the use of gases as heat transfer fluid inside the solar receivers. The success of this concept is related to the possibility to increase the fluid temperature above the actual maximum values: this requires the prediction of receiver efficiency, done on a physically consistent way by the model. An innovative integration between the solar field and an original gas turbine power plant modified in order to maximize thermal energy conversion, is presented. Index Terms Concentrated Solar Power, Heat transfer fluid, Parabolic-trough, Solar Energy. A I. INTRODUCTION mong renewable energy sources (RES), PT- CSP shows an undeniable "superiority", due to the high specific thermal power that the parabolas concentrate on receiver tubes which allows to have a high temperature operating working fluid. For these reasons, it demonstrates obvious advantages over other technologies that use solar energy. According to the actual technology, the fluid is used as high temperature source which feeds almost conventional thermoelectric power plants. As more the fluid temperature increases, as more the high temperature source assumes the entropy characteristics of fossil fuels. This gives to the PT-CSP technology the potentiality to replace, in a long term, fossil fuels: in the renewable energy-fossil fuels debate, the PT-CSP adds a new concept, and gets closer, for an engineering point of view, the inevitable energy transition closer from an engineering point of view [1]. However, a growing renewed interest in literature [2] has not been followed by an engineering and industrial development: this represents the weakest part of the CSP-PT technologies. An important breakthrough will be done by the Heat Transfer Fluid (HTF) whose allowable maximum temperature is a key factor to improve net electrical energy efficiency (and all the economic and financial parameters). Actual technology considers thermal oil as HTF. This use allows a limited exploitation of the solar radiation, limiting the maximum temperature to 450 C: overall efficiency is close to 25%. Greater values, but still limited by the maximum allowable higher temperature, are possible but they require the improvement of the conversion section. Even though these limits are recognized, almost the totality of the PT-CSP plants under construction makes use of this fluid, [3]. A step ahead will be done by molten salts as HTF [4]. Thanks to their chemical properties, they can reach maximum temperatures of 550 C, allowing a sensible efficiency plant improvement: they bring the thermoelectric conversion section toward the most advanced solutions. Some weakness points are still on the way: the solidification of the salt at temperature below than 290 C requires engineering care during operation. Energy storage systems are more critical in terms of reliability and efficiency; conventional alloys cannot be use for chemical compatibility cross coupled effects from a thermo-mechanical point of view prevent the use of available components (bellows, connectors between parabolas, deformable pipes, joints, sealing components, etc ). In spite of this difficulties, further research is under development with the aim to solve important engineering and operational aspects: the fluid sealing systems, the connection solution between parabolas and single receivers, as well as the improvement of solar receivers reliability, [5, 6]. All this elements will increase the financial interest toward this plants. Considering that a portfolio of technologies will be the right solution for PT-CSP plants, recently the use of gases as HTF has been considered, [7, 8]. Many components will take benefit of this choice including simplicity and HTF availability. Several operational aspects would be simplified, being more conventional with respect to the molten salt use (sealing, connection, material compatibility, etc. ). Many others will require additional scientific and technological attention. One limiting factor appears to be the operational pressure increase the Heat Collector Element (HCE). When using gases due to the need of keeping as possible the thermal capacity: the design of the receiver at very high pressure and temperature, is unconventional as well as devices to improve convective heat transfer between HTF and receiver. Nevertheless, gases will have a great future in the sector. In spite of this intense interest, the conversion sections have not received similar attention. They are the replica of conventional thermoelectric plants whose performances are limited by the lower value of the high temperature source (HTF instead of combustion gases) and by some choices which simplify the plants (when fed with CSP). DOI: / _2.1.30

2 Moreover a thermoelectric plant requires the vapor condensation which calls for a low temperature source. This represents a severe limitation on the PT-CSP sites which have their most suitable position in desert areas. It is not surprising that the most important large scale CSP plants [9], are placed in the coastal areas to take benefit from sea water. An interesting concept which improves profitability of these plants is represented by the possibility to manage CSP plants using sun during the day time and fossil fuels during the night: the plant modifications to allow for this double management are acceptable, even though the philosophy of the renewable generation looses a little bit of significance. A contribution to a system approach to CSP plants has been offered by the Authors recently, [10]. The result has been a comprehensive mathematical modeling which has as input the solar radiation and as output the temperature of the HTF. The model takes benefits from previous studies and privileges a global approach, [11]: all the design parameters of a CSP solar field can be setup and the HTF increase can be calculated, considering thermal oil, molten salt, gases. Thermal efficiency prediction is a key point when the receiver temperature is increased in order to favour higher HTF temperature. The paper preliminary outlined the possibility to used gas as HTF directly inside gas turbine plants, strongly simplifying the conversion section. In this paper the authors give further consistency to this concept, exploring the effects of: a) maximum operating HTF pressure; b) different HTF, having considered air and carbon dioxide; on the main parameters that are: (1) overall plant efficiency and (2) overall receiver length i.e. cost and reliability. The conversion section (fed by the solar fields) takes into consideration a modified unconventional gas turbine plant which approaches an Ericsson cycle. Thanks to sequences of inter-cooled compressions and inter-reheated expansions (fed by the solar fields), the overall plant efficiency demonstrates values close to the actual technologies (thermoelectric plants), in spite of the great simplicity which means cost reduction and reliability increase. The power plant which follows, when compared with conventional CSP-PT using thermal oil as HTF, opens new ways to a greater flexibility and to a number of not-only energetic applications. bellows are used in order to accounts for the different thermal expansion between the steel pipe and the glass envelope. HCE is radiated from the concentrate solar power due to sun radiation, which produces a series of processes which ultimately heat up the working fluid. They are: a) forced convection between glass tube and external air; b) radiation between the glass and the metallic tubes; c) forced convection between metallic tube and HTF. Inside glass and metallic tube, only conduction takes place which differentiate temperature at the corresponding inner and outer surface. The vacuum presents between glass and metallic tubes make almost nil the relative convection. Therefore, the solar receiver introduces five thermal states, being the HTF temperature the fifth. In steady state conduction, the thermal fluxes crossing the solar receiver must be equal. As HTF moves inside the receiver, its temperature increase and entrains toward higher temperature of the metallic tube and the glass tubes. Physical processes energy conservation equation produce a system of five not-linear algebraic equations where unknown are the five thermal states: thermal flux from the outside is specified. Appendix synthesize such equation; for a detailed description of the model, reference in [12] and [13] can be considered. Temperature variation along receiver length are stated by the energy conservation equation between the convective heat transfer reaching HTF and its temperature increase (Eq. A.1a-f in Appendix). Pressure drops must be taking into account, considering the overall great length of the receivers; when using gases, they can be calculated according the Eq. A.2a-c in Appendix. The most important property of the HCE is the thermal efficiency, defined as: Eq. 1 allows for the calculation of the HTF temperature increase when mass flow rate and c p are known as well as the solar radiation. (1) II. SOLAR RECEIVER MODELING In a CSP plant the key element is the one that collects heat increasing solar power, so the solar receiver. This consists of an absorber metallic tube surrounded by a glass tube. The inner tube is done in order to improve the energy collected: it combines a high absorbance for the solar radiation (short wavelengths) with low emissivity in the temperature range in which the surface emits radiation (long wavelengths). The external anti-reflective glass tube absorbs long wavelengths emitted by the metallic tube and protect its absorbent selective surface from oxidation due to vacuum between the two tubes. Compensation system made by III. ENERGY CONVERSION SECTION USING GAS TURBINES This paper considers a gas as working fluid which is expanded directly in gas turbines. Solar fields behave as a high temperature source while environment is considered as a low temperature source. From previous studies [14], carbon dioxide and air have been considered. When using carbon dioxide a heat exchanger as to be considered fed by external air; when using air it is simply discharged into atmosphere. So, HTF realizes a gas turbine cycle where compression, heating and expansion take place. In order to minimize compression work and maximize expansion work, sequences of inter-cooled compressions and inter-reheated expansions have been considered: in this way, real cycle approaches an

3 Ericsson cycle whose efficiency is equal to that of the Carnot cycle. This will require that regeneration takes place thanks to the high temperature value of the gas which leaves the last turbine and to the low temperature of the gas leaving the last compression (Discrete Ericsson Cycle - DEC). Making reference to three compression and expansion stages, the plant layout is in Figure 1: solar fields use is limited to the high temperature transformation once regeneration is included. Fig. 1. Gas turbine plant layout. When using air, the fluid is directly discharged into atmosphere. Contrary, when using carbon dioxide, a heat exchanger (represented in dotted line) has to be considered. In the figure, solar receivers are operated in parallel to limit maximum velocity inside them. In fact, proceeding inside the receiver, the gas is accelerated because of the temperature increase and pressure decrease. Mass conservation equation leads fluid velocity toward high values and these must be limited for noise and for limiting pressure losses. So, when the maximum allowable value was reached, working fluid flow rate was split in branches operated in parallel. When the mathematical model of the HCE and that of the thermodynamic transformation are integrated, temperature and pressure of the fluid inside the receiver as well as thermodynamic performances, of the conversion section can be calculated. By iterating the two models, the main items of the design of the CSP plant can be defined. IV. RESULTS In this work, a comparison between two plants with, respectively, air and CO 2 as HTF is considered. Figure 2 show the variations of HTF and metal temperature, HCE efficiency, absorbed power, outer fluid velocity long receiver length for three maximum pressure values: 2, 5, 10 MPa. Fig. 3. Variation of termo-fluodynamic parameter along the receiver length.

4 The asymptotic variation of HTF temperature showed versus length demonstrates that the receiver is progressively less efficient in order to transfer heat toward HTF: the flux radiated from the sun, in fact, remains constant (along receiver length) and the decreasing slope of the HTF temperature is the result of a greater heat loss. The temperature difference between HTF and metal decreases by increasing pressure of the gas, with opposite trend respect to convective heat transfer coefficient. The receiver efficiency, calculated as in Equation 1, progressively decreases as metal temperature increase and this contribute additionally to the overall receiver efficiency reduction. This is due to the increase of radiation and to the additional increase of the emissivity of the external coatings. Due to this, the thermal power received from HTF doesn t increase any more: all the thermal power that the receiver receive from the sun is re-radiated outside. Table I reports main geometrical data of the HCE: they refer to the actual technological standards [12]. TABLE I GEOMETRICAL DATA OF THE HCE Parabolic focus 1.61 m Parabolic reflector aperture 2.88 m Outer envelope diameter, m Inner envelope diameter, m Outer absorber diameter, m Inner absorber diameter, m Focus axis deviation m The performances of the CSP plant reported in Figure 1, have been calculated for a maximum pressure equal to 2 MPa, maximum temperature to K. relevant data are reported in Table II. TABLE II DEC PARAMETERS Direct irradiation 900 W/m 2 HTF inlet temperature K HTF inlet pressure 0.1 MPa Pinch point at the inter-cooling heat exchanger ΔT, 20 K Pinch point at the regenerator ΔT, 30 K Number of turbines 3 Number of compressors 3 Pressure ratio of each compressor β 2.7 Compressor (adiabatic) efficiency 0.85 Turbine (adiabatic) efficiency 0.89 Ericsson cycle efficiency (Carnot) 0.66 Air mass flow rate 1 kg/s CO 2 mass flow rate 1.4 kg/s For the two fields, the same velocity has been considered at the inlet of the first solar receiver, in order to compare also from a thermodynamic point of view its performances. This condition produces a different HTF flow rate as it is reported in Table II. Main reference points, characterizing the thermodynamic cycle, are reported in Figure 3: bolded lines represent the heating and the cooling of HTF inside the regenerator. The required overheatings happen inside solar receiver. Fig.3. DEC T-s plane. Table III puts in evidence the difference between air and CO 2. Specific work and cycle efficiency are respectively 287 kj/kg and 35% for CO 2, 274 kj/kg and 30% for air. The value of efficiency appears really interesting considering the simplicity of the plant. Similar values are today reached when a thermoelectric section is considered as conversion technology. TABLE III THERMODYNAMIC PROPERTIES * Temperature Pressure Enthalpy [K] [MPa] [kj/kg] Air CO 2 Air CO 2 Air CO *Adopted reference state: 0 kj/kg for enthalpy, 0 kj/kg-k for entropy, K for temperature and 0.1 MPa for pressure.

5 From a thermodynamic point of view, the use of CO 2 is more interesting in terms of cycle efficiency and overall power produced. As previously noted, when a fixed max velocity is reached inside HCE (20 m/s), flow rate is splitted in branches which operate in parallel. This condition produces that the first solar field is made by 3 branches in parallel, the second requires 8 branches and the 20 branches. This number remain equal independently from the fluid type. For each solar field, the overall length of the solar receiver (proportional to costs) doesn t remain the same, being it related to heat transfer conditions between fluid and metallic tube. For air, the respective length of each branch is: 33 m for the first field, 13 m for the second and 7 m for the third for total 335 m of receivers length. CO 2 fields require 49 m for the first solar field, 18 m and 9 m for the second and the third, respectively, for total 480 m. Figure 4 and 5 compare the overall receiver length, when using air or CO 2, as a function of the maximum pressure. If operating pressure is increased, the overall receiver length increases. This applies to the three solar fields even though in a different way between the three sections. The behavior is due to the fluid flow rate which increases when increasing pressure, requiring more solar radiation to reach a given fixed temperature. When the operating pressure increases, the receiver length is not proportional to the pressure ratio increase, even though the flow rate increases exactly in a proportional way. Figure 4 shows that for a pressure increase of a factor of two, in the first step (2-5 MPa) a very high length increase (mainly for the first solar field) is produced; this doesn t happen for the second pressure step (5-10 MPa). Similar results even though not so pronounced are for the second and third solar field: main reason of this behavior is receiver efficiency variation. The use of CO 2 keeps the same behavior obtained for the air and requires longer receiver length mainly due the fact that CO 2 flow rate is greater than air flow rate and when a fixed maximum temperature is fixed, higher heat must be exchanged toward the HCE. This would lead to a cost increase proportional with the receiver length. Fig. 4. Collectors length for the plant that use air as HTF. Fig. 5. Collectors length for the plant that use CO 2 as HTF. V. CONCLUSION Parabolic trough concentrating solar plants (PT-CSP) represent a very promising technology able to give to the solar energy as renewable source a new role: some intrinsic thermodynamic features allow to distinguish between its potentiality and those of the others renewable technologies. PT-CSP, in fact, produce a high temperature medium whose energy can be converted using actual technology into electricity. In a certain sense, PT-CSP offer a technological bridge among the conventional energy economy and an expected one based on renewables. One of the main limiting aspect is related to the high investment cost and weak profitability, even though the international interest is huge and networks which put together energy production, social development, desalination, territory valorization and mutual shared interests (for example, the Desertec plan) are more than a concept and proceed with an increasing international agreement. Increasing reliability and decreasing costs call for a plant simplification with respect to the existing technologies and also for a plant which is profitable event when reduced in size. The paper discusses the performances of a new energy conversion section based on gas turbine plants directly fed by the fluid which carries away, moving inside the receiver, the concentrated thermal power. The fluid considered in this paper are air and CO 2, both easily usable and available. The gas turbine plant is organized trough in a sequence of intercooled compressions and re-heated expansions in order to approach, after a recovery section among the gas leaving the last turbine and the itself leaving the last compressor, an Ericsson cycle. The remaining high temperature part of the plant is fed by solar fields. In real case, the compression and the expansion are not isothermal, so the cycle has been called DEC, Discrete Ericsson Cycle. The performances of the overall CSP-PT have

6 been calculated thanks to a comprehensive mathematical model which has in input the solar irradiation and the design aspects of the plants (parabolas, receiver type, DEC layout, type of fluid, base thermo fluid dynamic conditions - inlet pressure and temperature, velocity, maximum temperature, etc ): it calculates, among all the other parameters, the length of the solar fields and their layouts, having respected some important operational conditions (maximum velocity inside receivers, pinch points heat exchangers, maximum metal temperature, receiver efficiency control, etc ). CO 2 has several thermo fluid dynamic advantaged with respect to air when considered as working fluid: (1) smaller fluid velocity increase inside receivers; (2) greater receiver efficiency; (3) smaller temperature difference between fluid and metal. These positive aspects give a greater specific work and efficiency to DEC, demonstrating a thermodynamic superiority. From a plant point of view, the use of CO 2 requires longer receivers having fixed a maximum temperature of the cycle and the same minimum velocity of HCF when entering the first solar field: this would lead to a proportional cost receivers increase. APPENDIX,, 2,,,,,,,,,,,,,, 1 1,,,,,,,,,, 1 1,, 2,,,, 2,,,,,,,,, cos 4 (A.1 a ) (A.1 b ) (A.1 c ) (A.1 d ) (A.1 e ) 2 8 4, 2 0,184, 2,,,,, UNITS FOR FLUO-THERMODYNAMIC PROPERTIES Symbol Quantity heat flux per unit length convection heat transfer coefficient mass flow rate Nusselt number Prandtl number Reynold number Rayleigh number temperature diameter enthalpy kinematic viscosity density dynamic viscosity specific heat capacity at constant pressure mass flow rate thermal conductance pressure ratio velocity Stefan-Boltzmann constant emissivity absorptance transmittance efficiency incident angle modifier solar incident angle from the normal to the projected collector area parabolic reflector aperture pressure drop length friction factor H solar irradiation SUBSCRIPTS atmosphere Sky glass envelope absorber pipe solar fluid metal outer inner (A.1 f ) (A.2 a ) (A.2 b ) (A.2 c )

7 optical thermal Discrete Ericsson Cycle convention process conduction process radiation process ACKNOWLEDGMENT The activity has been done in the framework of two research programs. The first concerns unconventional sealing systems when using molten salt in PT-CSP plants by Meccanotecnica Umbra Group in Campello sul Clitunno, PG, Italy. The second deals about the improvement of receiver technology for oil, molten salt and gas as HTF by ASE, Archimede Solar Energy s.r.l. in Massa Martana, PG, Italy. The two Companies are very acknowledged. REFERENCES [1] R. Cipollone Renewable energy and new economic outlook 4 th Italian- Russian Forum: Cooperation for the Modernization and Innovation. Verona (Italy) October [2] A. Fernandez-Garcia et al. Parabolic-trough solar collectors and their applications. Elsevier, Renewable and Sustainable Energy Reviews 14 (2010) [3] CSP Today World Map, [4] M.J. Montes, A. Abánades, J. M. Martínez-Val. Thermofluidynamic Model and Comparative Analysis of Parabolic Trough Collectors Using Oil, Water/Steam, or Molten Salt as Heat Transfer Fluids ASME. Journal of Solar Energy Engineering. MAY 2010, Vol. 132 / [5] Meccanotecnica Umbra Group (various authors). Molten salt sealing systems for PT-CSP plants. Private communication. December [6] Archimede Solar Energy S.r.l. (various authors). Technological advancements for PT-CSP solar receiver. Private communication. June [7] C. Rubbia, Use of gas cooling in order to improve the reliability of the heat collection from the solar concentrators of ENEA design, [8] M-M. Rodrìguez-Garcìa et al., First experimental results of a solar PTC facility using gas as the heat transfer fluid, CIEMAT [9] L. E. Garcia Moreno, Concentrated Solar Power (CSP) in Desertec Analysis of technologies to secure and affordable energy supply. 6th IEEE International Conference, [10] R. Cipollone, A. Cinocca Integration between gas turbines and concentrated parabolic trough solar power plants. ASME 2012 International Mechanical Engineering Congress & Exposition, IMECE 2012, Houston, (TX). [11] R. Cipollone, A. Cinocca Integrazione tra turbine a gas e collettori parabolici lineari in impianti solari a concentrazione.. 67 ATI National Congress - Trieste (Italy) [12] R. Forristall. Heat transfer analysis and modeling of a parabolic trough solar receiver implemented in Engineering Equation Solver. Colorado: NREL-U.S Department of energy laboratory, [13] F. Burkholder and C. Kutscher Heat Loss Testing of Schott's 2008 PTR70 Parabolic Trough Receiver. Technical Report NREL/TP May [14] R. Cipollone, A. Cinocca Un modello matematico di supporto alle tecnologie del solare termodinamico a concentrazione.. 66 ATI National Congress - Cosenza (Italy) Roberto Cipollone is a full Professor in Interaction between Environment and thermal engines at the University of L Aquila. He graduated as Electric Engineer with honours in December 1980 at the University of L Aquila and in 1987 he got the Ph. D. degree after a period during which he joined the Von Karman for Fluid Dynamics (Brussels), winning the 1984 Belgian Government Prize. In 1988, he joined the Department of Energetics as Researcher, in 1991 as a Senior Researcher, in 1993 as Associate Professor, in 1994 as Full Professor. The scientific activity of Prof. Cipollone has treated many sectors around the thermodynamics applied to machines and thermal engines. Andrea Cinocca is a Ph.D. student in Energy technology and interaction with Environment at the Department of Industrial and Information Engineering and Economics, University of L Aquila (Italy). He received an M.S. degree in Environmental and Territory Engineering from University of L Aquila in 2008 and a Master s degree in Energy and Environment Markets from Alma Mater Studiorum - University of Bologna and Nomisma Energia in His research interests are environmental and energetic planning of a territory, renewable energy (in particular CSP technologies), interaction between environment and thermal engines.