- Email: [email protected]
biomass mini power plant
Applied Thermal Engineering xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng
Research Paper
Numerical simulation of a hybrid concentrated solar power/biomass mini power plant João Soares ⇑, Armando C. Oliveira University of Porto, Dept Mechanical Eng, Rua Dr Roberto Frias, 4200-465 Porto, Portugal CIENER-INEGI, Rua Dr Roberto Frias, 4200-465 Porto, Portugal
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
Simulation model of a hybrid
Concentrating Solar Power/Biomass mini power plant. Annual simulations were carried out for solar-only and hybrid modes. Electrical power generation stability is achieved with hybridisation. The system efficiency experiences a huge boost from 3% to 10% with hybridisation. Hybridisation significantly reduces, although not eliminating, the need of storage.
a r t i c l e
i n f o
Article history: Received 10 December 2015 Revised 23 June 2016 Accepted 29 June 2016 Available online xxxx Keywords: CSP Biomass Hybridisation Generation
a b s t r a c t Renewable energy sources such as solar energy are characterised by a high degree of intermittence, sometimes leading to inability to meet the demand of a power system. Hybridisation with more stable renewable sources, such as biomass, represents a resourceful way of meeting energy demands uninterruptedly. In this paper, a hybrid renewable electricity generation system is presented and modelled. The system relies on a combination of concentrating solar energy and biomass sources to drive an Organic Rankine Cycle. The solar field is constituted by parabolic trough collectors, and a biogas boiler is used as backup energy. The system was designed and a prototype will be installed in Tunis, in the framework of the REELCOOP project, co-funded by the EU. A computer model was developed with a combination of EBSILON and EES. Annual simulations were carried out for solar-only and hybrid modes. The system annual yield is significantly improved with hybridisation, increasing from 3.4 to 9.6%, with enhancement of SF and ORC efficiencies. Electrical generation stabilisation was achieved during the whole year with the fulfilment of ORC minimum requirements, contrarily to 1420 generation hours per year if the system relies solely on solar energy. On the other hand, hybridisation promoted energy excess mostly in the summer months, demonstrating that hybridisation significantly reduces, although not eradicating, the need for storage. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction
⇑ Corresponding author. E-mail address: [email protected] (J. Soares).
Renewable electricity generation systems are increasingly used as a means of reducing harmful emissions and also reducing operational costs, by comparison to the use of fossil fuels. However,
http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180 1359-4311/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
2
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
renewable energy sources such as solar energy are characterised by a high degree of intermittence, sometimes unpredictable. This constraint leads to inability to meet the demand of a power system. Hybridisation with more stable renewable sources, such as biomass, represents a resourceful way of meeting energy demands uninterruptedly. Besides stability, hybridisation with biomass allows a fully renewable solution, at the same time promoting security of energy supply. Concentrated Solar Power (CSP) plants require abundant solar radiation to be feasible and profitable [1], abutting the implementation to remote areas, far-off power consumption centres. Furthermore, the intermittent nature of solar radiation emphasises the generation stability drawback. To overcome such issues, usually CSP plants are designed in the range of 100 MWel [2], with intensive capital investment and financial risk, taking advantage from economy of scales. One of the main advantages of CSP over other renewable systems is the ability to provide dispatchable power, usually achieved through thermal energy storage (TES). Although energy dispatchability has been widely proven with TES, it is still a costly solution [3]. Facing a huge competition from other non-dispatchable renewable energy technologies (e.g. photovoltaics) [4], hybridisation presents a potential solution for forthcoming solar plants. Within the concept of fully renewable power systems, biomass is the ideal contender. The concept of CSP/biomass hybridisation relies on the ability of both systems to supply thermal energy in order to drive a power generation block. The advantages of this synergy go beyond dispatchability and renewable energy generation. They also include operation stability and flexibility, joint use of power plant equipment and associated cost reduction, as well as allowing CSP migration from desert areas to load centres [1,2,5–9]. Furthermore, during daylight periods, when electricity prices are usually higher, solar radiation is abundant and the system can run with larger solar shares resulting in a reduction of the levelised cost of energy. Despite the undeniable advantages of such systems, hybridisation still presents ambitious challenges, in order to guarantee that the system operates as a whole, and power reliability and stability are achieved. In this paper, a simulation model and results for a CSP/biomass hybrid mini power plant are presented. Annual simulations were carried out for solar-only and hybrid modes. Distinct operation ranges and boiler sizes were analysed. Simulation results are presented, such as: solar field annual generated heat and efficiency, boiler efficiency and biogas consumption, annual generated electrical energy and ORC efficiency, dumped heat, solar and biomass shares, and system global efficiency. Hourly results are presented for standard days, with and without hybridisation, showing the advantages of hybridisation. The solar field (SF) is constituted by parabolic trough collectors (PTC) with a net aperture area of 984 m2. As backup energy, a biogas boiler is used, fed through digestion of organic food waste. The nominal Organic Rankine Cycle (ORC) electrical output is 60 kW. The system was designed and a prototype will be installed in Tunis, in the framework of the REELCOOP project, co-funded by the EU [10]. 2. The hybrid mini power plant The hybrid mini power plant (Fig. 1) has a nominal electrical output of 60 kW, and relies on a regenerative ORC as power generation system, developed by Zuccato Energia. The turbine/generator block was adapted to assure operation at partial load, in order to compensate solar energy fluctuations, with a nominal gross efficiency reaching 13.3%. The ORC will be driven by saturated steam at 170 °C, which allows the power circuit to also operate
with available waste heat. Thermal generation will be achieved either from solar energy or biomass, or from the combination of both. The solar field relies on parabolic trough collector technology, and is constituted by 3 parallel loops of 4 PTMx/hp-36 collectors developed by Soltigua, with a net collecting surface of 984 m2. Direct Steam Generation (DSG) will be achieved in the solar field, and the recirculation concept was adopted. The solar field will be supplied with subcooled water and partial evaporation will take place in the solar collectors. The water/steam mixture is then separated in a steam drum, and therefore only saturated steam leaves the solar field. The leftover water is then recirculated. Complete evaporation enhances control complexity, implying unnecessary risks over solar the collector absorbers [11]. Auxiliary energy will be provided by a biogas steam boiler. The biogas will be produced by anaerobic digestion of canteen organic waste remains, showing a potential solution for the problem that waste disposal represents [12]. The system layout allows either hybrid or individual operation with each thermal source (solar-only or biogas-only). In order to reduce thermal energy waste as well as biogas consumption and to compensate short transients from solar power, a storage tank was foreseen in the project. Since the storage tank will be charged with saturated steam from the solar field, an isothermal latent thermal energy storage concept has been adopted. Whilst typical TES systems concern sensible heat storage with temperature change, the latent heat solution uses Phase Change Materials (PCM). 3. Simulation model The developed simulation model encompasses two stages. First, the solar field, the water/steam cycle and the boiler were analysed using a commercial software: EBSILONÒ Professional. This stage includes the simulation of the thermal generation system, for different operation profiles. The second stage concerns the power block circuit analysis that was carried out using EES software. The model required that relevant properties of the working fluid (Solvay SES36) were introduced, as they were not available either in the EES database nor in EBSILON. 3.1. Thermal generation system model EBSILONÒ Professional is a commercial simulation software designed for power plant and thermodynamic process analysis. The software includes a wide range of component libraries, for which steady state thermodynamic balances are simulated, under design and off-design conditions. The off-design calculations are carried out considering a linear relation between the off-design and design mass flow rate or/and thermodynamic properties (e.g. pressure, temperature, enthalpy). Simulations were carried out with an iteration precision of 107. In addition to the general EBSILON libraries, the EbsSolar, time-series, and EbsScript modules were used. The EbsSolar module consists of a component library (e.g. line focusing collectors, sun) specifically designed for solar thermal power plant simulations. To describe the solar power plant dynamic behaviour, the simulations were carried out, on an hour-by-hour basis, by using a time-series function. Furthermore, the detailed dynamic behaviour of the solar collectors and headers was considered. In real operation, the power plant is controlled depending on numerous external and internal variables. In order to describe this behaviour at different operation stages, a Pascal based code was developed using an EbsScript tool.
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
3
Fig. 1. Hybrid mini power plant layout [11].
Fig. 2. Thermal generation system simulation scheme.
The simplified system simulation scheme is presented in Fig. 2. The solar field model includes the sun, parabolic trough collectors, distributing and collecting headers and recirculating and feed water pumps. The biogas boiler was modelled throughout a heat injection component.
The model requires as input, metrological data (i.e. direct normal irradiation, ambient temperature). The data were obtained using Meteonorm software for the prototype site location (Tunis, Tunisia), on an hourly basis for a typical metrological year.
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
4
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
The Sun acts as interface between the meteorological data and incident radiation on the collectors. Sun angles, i.e. height (hs) and azimuth (cs) are calculated according to DIN-5034 standard, using as input the local time, latitude and longitude. In view of collector layout similarity, i.e. north-south orientation and non-slope, the normal irradiation incident angle (hi) and the collector tracking angle (q) can be determined [13],
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi cosðhi Þ ¼ 1 cos2 ðhs Þ cos2 ðcs Þ
ð1Þ
tan q ¼ sinðcs Þ= tanðhs Þ
ð2Þ
Therefore, the available beam irradiation at the collector surface (Ib), is a function of the incident angle (hi) and the direct normal irradiation (DNI),
Ib ¼ DNI cos hi
ð3Þ
For modelling the parabolic trough collectors, the line focusing solar collector component is used. The available energy within each collector (Qsolar) is defined as a function of the incident beam irradiation (Ib) at the net collector aperture area (Anet), considering the nominal collector optical efficiency (gopt), incident angle modifier (IAM), collector focus state (f), shading losses (gshad), and reflecting mirror cleanliness (gclean) [14],
Q solar ¼ Ib Anet gopt IAM f gshad gclean
ð4Þ
Manufacturer information concerning the nominal optical efficiency and incident angle modifier (IAM) were used for Soltigua’s collectors [15]. The IAM correction is a function of the incident angle, determined by a second order interpolation of table based values. When the sun height is low (e.g. sun appearing in the horizon), parallel collector row shading is expected to occur. This effect is determined by the collector row distance (lrow) and aperture width (Awidth) as well as the tracking angle (q) [14],
gshad ¼ 1 min ½1; max ð0; 1 lrow cos q=Awidth Þ
ð5Þ
A constant yearly mirror cleanliness factor (gclean) of 0.97 was considered [16]. The effective heat received by the heat transfer fluid (Qeff) is defined as the energy balance between the available solar energy (Qsolar) and thermal losses (Qloss). The heat losses at the collectors result from the temperature difference between the heat transfer fluid and the ambient air. A linear heat loss coefficient was considered according to manufacturer’s data [15]. The direct steam generation concept implies water evaporation within the collectors. Therefore, pressure losses (DPcoll) were calculated considering a two-phase flow model. The model entails the calculation of pressure losses considering a smooth pipe (DPcoll,S) and rough pipe (DPcoll,R). Afterwards, the worst case is considered:
DPcoll ¼ maxðDPcoll;S ; DPcoll;R Þ
ð6Þ
The smooth pipe calculations rely on the Friedel two-phase flow pressure drop model [17]. The model consists in the calculation of the pressure drop in the pipe as if it was a single-phase flow, both liquid and vapour. Subsequently, a correction factor is used, calculated from the Weber and Froude number for the liquid-phase flow as well as the steam quality within the pipe. On the other hand, the aforementioned model discards the pipe roughness. Thus, a correlation between single-phase and two-phase flow pressure drop in rough pipes is used. Initially, the pressure drop is calculated considering single-phase flow, for _ liquid (DPcoll,R,L) and gas (DPcoll,R,G), using the mass flow rate (m), pipe internal diameter (Dint), fluid density (qfluid) and a friction factor (f) [18],
DPcoll;RL
or G
¼ fL
or G
_2 m 2Dint qfluid
ð7Þ L or G
The friction factor (f) is a function of the pipe internal diameter (Dint), pipe roughness (e) as well as the Reynolds number (Re). In order to calculate f directly, the Swamee–Jain equation [19] is used,
fL
or G
2 e 5:74 ¼ 0:25 log10 þ 0:9 3:7Dint Re
ð8Þ
Thereafter the pressure drop in the absorber tube is calculated as a function of the steam quality (X),
DPcoll;R ¼ DPcoll;RL þ X ðDP coll;RG DPcoll;RL Þ
ð9Þ
In each case, the model uses the thermodynamics properties (e.g. enthalpy, pressure, steam quality) at the collector centre. For simulation purposes, a one-loop simplification was considered, in order to reduce the computational effort. The other two loops were accounted through the use of distributing and collecting headers. These components allow considering the fluid stream distribution and collection, without neglecting mass flow, heat and momentum balances. Whilst mass flow rate within the loops is considered identical, enthalpy and pressure are different due to heat and pressure losses along the header. The header pressure loss model is identical to the one for solar collectors, with a single-phase flow consideration in the distributing header and a two-phase flow in the collecting header. As inputs, the header length (60 m) and a constant diameter were considered, divided into three identical sections. The internal diameters of the distributing and collecting headers, were defined as 0.0418 and 0.053 m, respectively. Heat losses (Qloss,H) within the header are calculated by the radial heat conduction in the insulting material [20],
Q loss;H ¼ 2pkins
T T amb fluid ln Dext;ins =Dint;ins
ð10Þ
Calculations are carried out for each section, with fluid temperature (Tfluid) determined at the section centre. The insulation thermal conductivity (kins) was defined as 0.1 W/mK and the ratio between the outer (Dext,ins) and inner (Dext,ins) diameters as 2.24. Heat and pressure losses in the pipe connections within the loop are considered as well. Pressure drop is calculated by the application of Bernoulli’s theorem for a specific pipe geometry. Heat losses are calculated using Eq. (10). The transient behaviour of the solar field is modelled through the use of an indirect storage (IS) component. The implementation of an IS component was restricted to each of the solar collectors and both headers, where the highest temperature gradients are expected to occur. The IS model calculates the transient heat exchange in a pipe between the working fluid (i.e. water/steam) and either the collector absorber tube or piping. The Fourier heat transfer differential equation [20],
@T @ 2 T pipe @ 2 T pipe qcp pipe ¼ k þ @t @x2 @y2
!
ð11Þ
is discretised in a two-dimensional space domain using a finite volume method, and in time by an iterative Crank-Nicholson method [21], The convective heat flux (Q_ pipe;fluid ) between the pipe walls and the heat transfer fluid is calculated using the Newton’s cooling law [20], for each control volume area (A) and assuming a constant heat transfer coefficient along the pipe,
Q_ pipe;fluid ¼ aAðT pipe;wall T fluid Þ
ð12Þ
The dynamic behaviour of the fluid temperature along the pipe, is modelled by the energy and mass balance:
_ in hout Þ mfluid cp;fluid mðh
@T fluid ¼ Q_ pipe;fluid @t
ð13Þ
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
A higher number of elements in the mesh, significantly improves simulation accuracy, however at the expense of increased computational effort. Based on a sensitivity analysis and considering that the temperature gradient is much higher in the axial direction, the control volume was divided into 60 elements (i.e. 30 in the axial direction and 2 in the radial direction). The time-step (Dt) was defined considering the Courant–Friedrichs–Le wy stability condition [22], for each element size (Dx), using the fluid velocity (Vfluid) at the pipe centre,
V fluid Dt=Dx < 1
ð14Þ
The nominal convective heat transfer coefficient (a) used in Eq. (12), was calculated for each collector and header,
a ¼ kNuD =Dint
ð15Þ
The Nusselt number (NuD) was estimated by the Dittus-Boelter equation, using the Reynolds (Re) and Prandtl (Pr) numbers,
NuD ¼ 0:023Re0:8 Pr 0:4
ð16Þ
3.1.1. Hybridisation Hybridisation was modelled through a heat injection component, representative of the boiler. This component acts as an ideal heat exchanger, promoting the interface between the boiler output and the mass flow rate of water/steam. In order to control the boiler output, a code was created using EbScript to impose operating limits according to the manufacturer data, as well as to calculate the required heat flow. The boiler output is controlled by the water/steam mass flow rate, acting as an auxiliary heater of the solar field. For simulation purposes, the Viessmann VITOMAX 200-HS boiler model with economiser was used. Two different boiler sizes were the object of analysis, with nominal heat outputs of 380 kWth (0.5 ton/h of saturated steam) and 530 kWth (0.7 ton/h of saturated steam), and both with a minimum thermal output of 100 kWth. In order to account for the transient behaviour of the boiler an IS component was used. As inputs, the water volume of the boiler was used and the mass of steel estimated, considering that the boiler takes half an hour from cold start to design conditions. For estimating biogas consumption, an additional computer model was created using Ebsilon, mainly constituted by a combustion chamber and two heat exchangers (i.e. steam evaporator and economiser) (see Fig. 3). The model uses the water/steam mass flow rate, as well as the inlet and outlet enthalpies to calculate the required boiler output. These inputs are obtained from the thermal generation system model results. The boiler efficiency, combustion heat output and flue gas temperature are determined as a function of the ratio between the boiler output and the rated output, considering the
5
manufacturer data. In the combustion chamber, the combustion of the mix of air and biogas is modelled, considering a 3% oxygen excess in the flue gas. As output it retrieves the biogas mass flow rate. Biogas consumption was estimated considering a yearly constant low heating value (19.27 MJ/m3). A study was conducted at ENIT, for estimating biogas production with the local canteen residues. The estimated value is about 60 m3/day. 3.2. Power block The regenerative Organic Rankine Cycle was simulated using EES, since property data of the organic fluid (Solvay SES36) were not available neither in the EES nor in the Ebsilon databases. The data values were introduced into tables from which the thermodynamic properties (e.g. enthalpy, entropy) are calculated by linear interpolation. It was assumed that the ORC operates at steady-state, with thermal inertia of the power block neglected. Furthermore, the evaporation and condensation temperatures are considered constant, as well as the pump efficiency and the temperature difference in the regenerator. As input the code requires the saturated steam mass flow rate obtained from the thermal generation system simulation results. The steam mass flow rate is used to calculate the turbine isentropic efficiency, by means of a second order equation based on manufacturer data. Within the ORC model, calculations for individual components (i.e. turbine, pump and heat exchangers) are achieved by applying the first law of thermodynamics. Irreversibilities are considered at the turbine and pump. As main output results the code provides the gross and net electrical power, organic fluid mass flow rate, as well as the parasitic consumption and the condenser thermal requirements. 3.3. Operation modes and control In order to optimise the system operation and to enhance similarity with real operation, the thermal generation system model is controlled and also distinct operation profiles are considered. EBSILON contains a wide-range of components suitable for control purposes, such as property value transmitters, calculators and simple linear controllers. Nevertheless, more complex control tasks, such as interchange between operation profiles, require the development and implementation of a code. Therefore, a computational code was developed, using a Pascal based language and implemented in the numerical model using the EbScript tool. 3.3.1. Solar-only operation mode The system steam production will be controlled by the mass flow balance at the steam drum. The same concept was used in the simulation model. During system operation the recirculating
Fig. 3. Boiler simulation scheme.
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
6
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
mass flow rate is kept constant at about 0.5 kg/s. In real conditions the feed water pump will operate when the water level in the steam drum drops below a predefined level. Since the simulations were carried out on an hourly basis, it was assumed that the feed water mass flow rate should balance the water removed from the recirculating pump, and so is intrinsically related to the steam quality after the collecting header. If the saturated steam mass flow rate that leaves the solar field exceeds the maximum requirements of the ORC, it is then separated using a simple splitter (see Fig. 2). In real operation, the solar collectors should change their state to partial defocused in order to control the steam production. In the simulation model, the excess of energy is accounted as dumped, providing an input for a PCM storage tank design. The last separation occurs before the ORC, through a water-steam drainer. This component is mostly used in the warm-up and cool-down profiles, to establish a more realistic thermodynamic balance. To obtain a more accurate approach in the simulation model, distinct operating profiles were created. A code was developed using EbScript to allow the automatic interchange between the profiles using dynamic variables, e.g. direct normal irradiation, hour of the day, mass flow rates, etc. The solar-only operating mode is constituted by four profiles: warm-up, operating, cool-down and stop. At the beginning of the day if DNI exceeds 200 W/m2 the collectors change their state to focus and the recirculating pump is activated. This represents the warm-up profile. The operation profile starts with solar field steam production, and operates as described before. At the end of the day, if DNI is less than 200 W/m2, the collectors change their state to defocus and water circulates until the system cools-down. At night, the system is off, with solar field collectors defocused and the pumps shut-down. Even during the night, the thermal inertia of the system as well as the heat losses to the ambient were considered. 3.3.2. Hybrid operation mode The hybrid operation analysis was carried out based on the assumptions of the system running 12 or 24 h daily, at ORC minimum and nominal power. Concerning the control, the hybrid mode comprehends four and two operating profiles for the 12 and 24 h regimes, respectively. The 12 h operating regime differs from the solar-only, on the warm-up and operating profile. The warm-up begins at 7:00 with the start of the boiler in order to warm-up the ORC and the feed water pipe of the solar field. If DNI exceeds 200 W/m2 the recirculating pump is activated and the solar collectors focused. During the operation regime (08:00–20:00) the hybrid mode is activated, with the boiler compensating the requirements (nominal or minimum) to drive the ORC turbine. This control is achieved by saturated steam flow rate balance. In preliminary simulation results it was noticed that during summer the system could start earlier (at 7:00), and the 12-h operation regime acted as a constraint to the solar field. To overcome this issue, in the beginning of the day if DNI is above 200 W/m2 the collectors are focused earlier. The 24-h regime just encompasses two operation modes, hybrid and boiler-only. During the daylight period the system operates in hybrid mode, and at night the collectors are defocused and the solar field is cooled down. After that the system relies solely on the boiler. For both cases the boiler minimum power of 100 kWth was considered, in order to assure electricity generation stability, during the predefined time operating range. Otherwise, the boiler would be submitted to consecutive start-ups and shutdowns, and shortages in the electrical generation would be expected, due to solar radiation transients.
4. Results If the system relies solely on solar energy the annual heat generated is about 663 MWhth. Almost one quarter of the heat is dumped, mostly related with energy dearth. The dumping rate results are divided in two items (excess and scarcity), representing the heat dumped due to the excess of energy or due to insufficient energy to drive the ORC turbine, respectively. The annual power generated is 61 MWhel with an ORC average annual efficiency of 9.2% and an annual running time of 1420 h. The system efficiency is 3.4% hindered by the excessive dumping rates (see Table 1). Hybridisation improved the solar field output by 3% (Table 2). This outcome is related with the system start-up, since the SF feed water is already warmer, and consequently less solar energy is required to achieve steam generation. Furthermore, this improvement is extended to the solar field annual efficiency. The second improvement of hybridisation was the extinction of dumping rates associated with scarcity of energy. The scarcity of energy was surpassed with hybridisation, with the fulfilment of the ORC minimum thermal power requirements. Moreover, ORC operation near nominal conditions enhanced the ORC efficiency. On the other hand, the excess of energy increased. This fact is related with minimum operating conditions of the boiler (100 kWth), leading to energy waste predominantly in the summer months, when solar radiation is highly available. The excess of energy can be reduced by implementing a storage tank in the system. The benefits extend beyond the ability to store the excess of solar field thermal energy. If the storage can provide more than 30 min of thermal energy requirements to drive the ORC turbine, the need of having the boiler in permanent operation is eliminated. In other words, it can act as system buffer in order to compensate thermal output fluctuations from the solar field and boiler, reducing the amount of wasted biogas. Despite the discontinuous operation of the boiler, due to the solar irradiance transients, the average biogas boiler efficiency is still high (about 93%) for all cases. As expected, electric generation is significantly increased with hybridisation. If the system depends only on solar energy, generation is confined to sunnier months. In spite of the favourable Tunisian climate, solar radiation monthly values show a noteworthy discrepancy between summer and winter. To enable electrical generation in low radiation months, the solar multiple should be increased, leading to costly and oversized systems. On the other hand, even system hybridisation for 12 h/day at ORC minimum power showed an annual generation of 167 MWhel, with a solar share of 44%. System annual efficiency experienced a huge boost with hybridisation, from 3.4% to almost 10%, due to the high efficiency boiler, along with improved efficiencies of the SF and ORC
Table 1 Solar-only simulation annual results. Direct Normal Irradiance – DNI Annual Heat Generated S.F. Specific Thermal Field Output Mean Annual SF Efficiency Dumping Rate – Excess Dumping Rate – Scarcity Annual Useful Heat – ORC Annual Power Generated Mean Annual ORC Efficiency ORC – Number of hours running Annual Dissipated Heat – Condenser Maximum Heat Dissipated – Condenser Mean Annual System Efficiency
1799.4 663 674 37.4% 7.9% 16.3% 503 61 9.2% 1420 437 387 3.4%
kWh/(m2 a) MWhth kWhth/m2 % % % MWhth MWhel % h MWhth kWhth %
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
7
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx Table 2 Hybrid simulation annual results. Hybrid – 24 h operation
Direct Normal Irradiance – DNI Annual Heat Generated S.F. Specific Thermal Field Output Mean Annual SF Efficiency Annual Heat Generated – Boiler Annual Combustion Heat – Boiler Mean Annual Boiler Efficiency Annual Biogas Consumption Average Biogas Consumption Solar Share Dumping Rate – Excess Dumping Rate – Scarcity Annual Useful Heat – ORC Annual Power Generated Mean Annual ORC Efficiency ORC – Number of hours running Annual Dissipated Heat – Condenser Maximum Heat Dissipated – Condenser Mean Annual System Efficiency
Hybrid – 12 h operation
530 kWth
380 kWth
ORC minimum
530 kWth
380 kWth
ORC minimum
Boiler
Boiler
Power
Boiler
Boiler
Power
1799.4 683 694 38.6% 3359 3619 92.8% 655 1795 17% 2.5% 0.0% 3941 515 12.7% 8760 3389 387 9.6%
1799.4 683 694 38.6% 2546 2746 92.7% 497 1362 21% 3.1% 0.0% 3128 380 11.8% 8760 2720 387 8.4%
1799.4 683 694 38.6% 1788 1918 93.2% 347 951 28% 4.1% 0.0% 2370 261 10.6% 8760 2089 387 7.1%
1799.4 683 694 38.6% 1425 1534 92.9% 278 761 32% 6.5% 0.5% 1971 257 12.2% 4380 1695 387 7.8%
1799.4 683 694 38.6% 1155 1244 92.8% 225 617 37% 7.5% 0.5% 1700 213 11.6% 4380 1472 387 7.1%
1799.4 683 694 38.6% 865 928 93.2% 168 460 44% 8.9% 0.0% 1410 167 10.8% 4380 1231 387 6.2%
kWh/(m2 a) MWhth kWhth/m2 % MWhth MWhth % dam3 m3/day % % % MWhth MWhel % h MWhth kWhth %
should be used only during start-up and shut-down, reducing the amount of biogas consumption and heat dumped. The boiler starts to operate at 07:00, at minimum power in order to warm-up the SF and ORC. From 8:00 to 20:00 nominal electricity generation is achieved mostly from solar energy. Either Summer or Winter solstice results, showed that electricity stabilisation can be achieved during the daylight period through hybridisation, moreover with a significant solar share in the summer (75%) and less relevant (13%) in the winter. This synergy is noteworthy, since during the daylight period electricity prices are usually higher, related to peaks in network consumption. The estimated biogas production of 60 m3/day is far below the consumption results. On an annual basis the excess of energy related with the boiler minimum operating conditions represents 738 h of operation and 14 dam3 per year. If we consider as an example the case of 12 h of operation at minimum power, this represents 9% of the annual biogas consumption and 16% of the running hours. This denotes a minor contribution to CSP/biomass hybridisation. Despite hybridisation allowing to relocate a power plant near urban centres where organic wastes are more abundant,
(Fig. 4). In the REELCOOP framework a 10% system efficiency was proposed as target, which is nearly attainable with hybridisation. One of the main advantages of hybridisation is the stability of the system, in order to promote dispatchability. This can be observed in the simulation for the 21st of December (Fig. 5), where the generator is operating 24 h at nominal power with the 530 kWth boiler. On the 21st of December, the boiler is supplying 450 kWhth until 9:00, fulfilling the ORC requirements during night operation. At this time the DNI is above 200 W/m2, and the collectors change their sate to focus. Steam production from the solar field starts at 10:00, with a thermal output of about 50 kWth. When compared with solar-only, the steam generation starts one hour later and with half of the hybrid production. During the day the solar field is unable to supply the minimum conditions to drive the ORC turbine. Nevertheless, it contributes to a reduction in the amount of required boiler energy and biogas consumption. On the Summer solstice (21st June), the generation stability was attained even with the smaller boiler (Fig. 6). However, the heat dumped due to excess of energy increased. In such days, the boiler Annual Power Generated [MWhel] Proposed Target Efficiency [%]
Mean Annual System Efficiency [%] 10%
10% 8.4% 7.8% 7.1%
7.1% 6.2%
3.4%
9.6% 515
380 257
261
530 kWth Boiler
ORC minimum Power
213 167
61
Solar Only
ORC minimum Power
380 kWth Boiler
12 hours
Hybrid - NP Boiler 1
530 kWth Boiler
24 hours
Fig. 4. Annual power generated and mean annual efficiency.
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
8
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
Solar Field Thermal Output -Hybrid
NET Power ORC
Boiler Thermal Output
Solar Field Thermal Output - Solar Only
500
900
450
800
400
700
350
600
300
500
250 400
200
DNI [W/m2]
Thermal and Electric Output [kW]
DNI
300
150 100
200
50
100
0
0 0
2
4
6
8
10
12
14
16
18
20
22
Hours
Solar Field Thermal Output
Boiler Thermal Output
NET Power ORC
DNI
500
900
450
800
400
700
350
600
300
500
250 400
200
DNI [W/m2]
Thermal and Electric Output [kW]
Fig. 5. Winter solstice 24-h hybrid operation with a 530 kWth boiler.
300
150 100
200
50
100
0
0 0
2
4
6
8
10
12
14
16
18
20
22
Hours Fig. 6. Summer solstice 12-h hybrid operation with a 380 kWth boiler.
500
Mean Daily Heat Dumped - Excess [kWhth]
Mean Daily Heat Dumped - Scarcity [kWhth] 435
450 400
371 325
350
332
325
300
322
273 239
250
208
184
200
224
231 202
150 76
76
48 0
130
121
100 50
382
370 342
36
17 January
10 February
March
April
May
June
July
August September October November December
Fig. 7. Heat dumped in the solar-only operation mode.
the absence of a well-established biomass market represents a drawback to the dissemination of these systems [1]. As aforementioned, either solar-only or hybrid operation results showed excess of thermal energy, mostly in the summer months
when solar radiation is abundant. The energy waste can be overcome with a storage tank. To define the ideal storage capacity the daily average values of the heat dumped due to energy excess were analysed for the solar-only mode (Fig. 7). The analysis was
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
9
not extended to the hybrid mode, since it enhances the dumped heat and consequently the storage capacity. Furthermore, this overestimation can easily be eliminated with a small storage capacity. The maximum value for the average daily dumped heat is 332 kWhth (June) and the minimum is 10 kWhth (December). It is worth to note that the maximum daily value of the excess heat does not allow to drive the ORC at nominal power for one hour.
grant agreement no. 608466 (www.reelcoop.com). All other partners involved in the development of Prototype 3 are greatly acknowledged: German Aerospace Centre (DLR, Germany), CIEMAT (Spain), ENIT (Tunisia), Soltigua (Italy), Zuccato Energia (Italy), AES (Tunisia).
5. Conclusion
[1] A. Colmenar-Santos, J.-L. Bonilla-Gómez, D. Borge-Diez, M. Castro-Gil, Hybridization of concentrated solar power plants with biogas production systems as an alternative to premiums: the case of Spain, Renew. Sustain. Energy Rev. 47 (2015) 186–197. [2] H.G. Jin, H. Hong, 12 - Hybridization of concentrating solar power (CSP) with fossil fuel power plants, in: K. Lovegrove, W. Stein (Eds.), Concentrating Solar Power Technology, Woodhead Publishing, 2012, pp. 395–420. [3] B. Coelho, S. Varga, A. Oliveira, A. Mendes, Optimization of an atmospheric air volumetric central receiver system: impact of solar multiple, storage capacity and control strategy, Renewable Energy 63 (2014) 392–401. [4] G.K. Singh, Solar power generation by PV (photovoltaic) technology: a review, Energy 53 (2013) 1–13. [5] A.A.A. Cot, J. Vall-Llovera, J. Aguiló, J.M. Arqué, Termosolar Borges: A Thermosolar Hybrid Plant with Biomass, in: Third International Symposium on Energy from Biomass and Waste, CISA, Environmental Sanitary Engineering Centre, Italy, Venice, Italy, 2010. [6] Á. Moreno-Pérez, P. Castellote-Olmo, Solar Parabolic Trough – Biomass Hybrid Plants: Features and Drawbacks, in: International Conference on Concentrating Solar Power and Chemical Energy Systems, SolarPACES 2010, Perpignan, France, 2010. [7] B. Coelho, P. Schwarzbözl, A. Oliveira, A. Mendes, Biomass and central receiver system (CRS) hybridization: volumetric air CRS and integration of a biomass waste direct burning boiler on steam cycle, Sol. Energy 86 (2012) 2912–2922. [8] J.H. Peterseim, S. White, A. Tadros, U. Hellwig, Concentrated solar power hybrid plants, which technologies are best suited for hybridisation?, Renewable Energy 57 (2013) 520–532 [9] J.H. Peterseim, S. White, A. Tadros, U. Hellwig, Concentrating solar power hybrid plants – enabling cost effective synergies, Renewable Energy 67 (2014) 178–185. [10] REELCOOP, REELCOOP project, in, 2015. [11] D. Krüger, A. Kenissi, S. Dieckmann, C. Bouden, A. Baba, A. Oliveira, J. Soares, E. R. Bravo, R.B. Cheikh, F. Orioli, D. Gasperini, K. Hennecke, H. Schenk, Pre-design of a mini CSP plant, Energy Procedia 69 (2015) 1613–1622. [12] A.C. Oliveira, B. Coelho, REELCOOP Project: Developing Renewable Energy Technologies for Electricity Generation, in: 12th International Conference on Sustainable Energy Technologies (SET2013), Hong Kong, 2013. [13] W.B. Stine, R.W. Harrigan, Solar Energy Fundamentals and Design: With Computer Applications, John Wiley & Sons, Incorporated, 1985. [14] STEAG Energy Services GmbH, Line focusing Solar Collector, in: Solar Library EbsSolar, 2015. [15] Soltigua, PTMx/hp Parabolic Trough Collector – Technical Data Sheet, in, 2013. [16] I. Llorente García, J.L. Álvarez, D. Blanco, Performance model for parabolic trough solar thermal power plants with thermal storage: comparison to operating plant data, Sol. Energy 85 (2011) 2443–2460. [17] H. Schmidt, A. Wellenhofer, S. Muschelknautz, J. Schmidt, F. Schmidt, D. Mewes, A. Mersmann, J. Stichlmair, L2 Two-Phase Gas-Liquid Flow, in: VDI Heat Atlas, Springer, Berlin, Heidelberg, 2010, pp. 1117–1180. [18] B.R. Munson, A.P. Rothmayer, T.H. Okiishi, Pre-design of a Mini CSP Plant, seventh ed., John Wiley & Sons Incorporated, 2012. [19] P.K. Swamee, A.K. Jain, Explicit equations for pipe flow problems, J. Hydraul. Eng. 102 (1976) 657–664. [20] F.P. Incropera, T.L. Bergman, D.P. DeWitt, A.S. Lavine, Fundamentals of Heat and Mass Transfer, Wiley, 2013. [21] R. Pawellek, S. Pulyaevrank, T. Hirsch, Transient simulation of a parabolic trough collector in EBSILON, in: International Conference on Concentrating Solar Power and Chemical Energy Systems, SolarPACES 2012, Marrakesh, Morocco, 2012. [22] R. Courant, K. Friedrichs, H. Lewy, Über die partiellen Differenzengleichungen der mathematischen Physik, Math. Ann. 100 (1928) 32–74.
In this paper, a simulation model and results of a hybrid CSP/biomass mini power plant were presented. The simulations were carried out on an hour-by-hour basis for a typical meteorological year, considering different scenarios regarding operating regimes and boiler sizes. If power generation is exclusively dependent on the solar field, electrical generation is mostly confined to sunnier months and negligible in the winter. The annual heat dumping rate ascends to almost one quarter, mostly related with the inability of the solar field to supply the minimum thermal energy to drive the ORC turbine. System hybridisation proved to stabilise the system regarding electrical power generation during the whole year. Additionally, the downside of the dumped heat, due to scarcity of thermal energy, was surpassed. Hybridisation improvements where extended as well to the SF and ORC efficiencies. The SF efficiency increased 3% since the system is already warmer in the morning, and solar radiation is used exclusively for steam generation. The ORC efficiency increase is in the range of 15 to 38%, and was achieved by a stable operation near turbine design conditions. On the other hand, there was an increase in the system energy excess due to the downsides of boiler start-up time (about 30 min) and minimum operational heat input (100 kWth). The simulated operation profiles were created on the basis of energy dispatchability and demand response ability. Therewith, boiler operation was extended to the system operation range to compensate possible short transients from solar power. This issue can be overcome with the implementation of a storage tank. From simulation results, the maximum average daily dumped heat by excess ascends to 332 kWhth. This value is quite small when compared with the system size. For example, it does not allow to run the system for one hour at nominal power. This proves that the solar field design is appropriate and hybridisation significantly reduces (although not totally eliminating) the need of storage. The improvements in the SF and ORC efficiencies, along with extended operation ranges and a highly efficient boiler, lead to a huge boost on system efficiency, which increases from 3.4 to 9.6%. As future work, system simulations will include the implementation of a PCM storage tank. More operation strategies, biomass resources, different climatic regions and a cost analysis will be addressed. A prototype of the system was developed and will be installed in Tunisia, for testing during 2016/17, allowing a validation of the simulation results.
References
Acknowledgements The REELCOOP project receives funding from the European Union Seventh Framework Programme (FP7/2007-2013), under
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
Contents lists available at ScienceDirect
Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng
Research Paper
Numerical simulation of a hybrid concentrated solar power/biomass mini power plant João Soares ⇑, Armando C. Oliveira University of Porto, Dept Mechanical Eng, Rua Dr Roberto Frias, 4200-465 Porto, Portugal CIENER-INEGI, Rua Dr Roberto Frias, 4200-465 Porto, Portugal
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
Simulation model of a hybrid
Concentrating Solar Power/Biomass mini power plant. Annual simulations were carried out for solar-only and hybrid modes. Electrical power generation stability is achieved with hybridisation. The system efficiency experiences a huge boost from 3% to 10% with hybridisation. Hybridisation significantly reduces, although not eliminating, the need of storage.
a r t i c l e
i n f o
Article history: Received 10 December 2015 Revised 23 June 2016 Accepted 29 June 2016 Available online xxxx Keywords: CSP Biomass Hybridisation Generation
a b s t r a c t Renewable energy sources such as solar energy are characterised by a high degree of intermittence, sometimes leading to inability to meet the demand of a power system. Hybridisation with more stable renewable sources, such as biomass, represents a resourceful way of meeting energy demands uninterruptedly. In this paper, a hybrid renewable electricity generation system is presented and modelled. The system relies on a combination of concentrating solar energy and biomass sources to drive an Organic Rankine Cycle. The solar field is constituted by parabolic trough collectors, and a biogas boiler is used as backup energy. The system was designed and a prototype will be installed in Tunis, in the framework of the REELCOOP project, co-funded by the EU. A computer model was developed with a combination of EBSILON and EES. Annual simulations were carried out for solar-only and hybrid modes. The system annual yield is significantly improved with hybridisation, increasing from 3.4 to 9.6%, with enhancement of SF and ORC efficiencies. Electrical generation stabilisation was achieved during the whole year with the fulfilment of ORC minimum requirements, contrarily to 1420 generation hours per year if the system relies solely on solar energy. On the other hand, hybridisation promoted energy excess mostly in the summer months, demonstrating that hybridisation significantly reduces, although not eradicating, the need for storage. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction
⇑ Corresponding author. E-mail address: [email protected] (J. Soares).
Renewable electricity generation systems are increasingly used as a means of reducing harmful emissions and also reducing operational costs, by comparison to the use of fossil fuels. However,
http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180 1359-4311/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
2
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
renewable energy sources such as solar energy are characterised by a high degree of intermittence, sometimes unpredictable. This constraint leads to inability to meet the demand of a power system. Hybridisation with more stable renewable sources, such as biomass, represents a resourceful way of meeting energy demands uninterruptedly. Besides stability, hybridisation with biomass allows a fully renewable solution, at the same time promoting security of energy supply. Concentrated Solar Power (CSP) plants require abundant solar radiation to be feasible and profitable [1], abutting the implementation to remote areas, far-off power consumption centres. Furthermore, the intermittent nature of solar radiation emphasises the generation stability drawback. To overcome such issues, usually CSP plants are designed in the range of 100 MWel [2], with intensive capital investment and financial risk, taking advantage from economy of scales. One of the main advantages of CSP over other renewable systems is the ability to provide dispatchable power, usually achieved through thermal energy storage (TES). Although energy dispatchability has been widely proven with TES, it is still a costly solution [3]. Facing a huge competition from other non-dispatchable renewable energy technologies (e.g. photovoltaics) [4], hybridisation presents a potential solution for forthcoming solar plants. Within the concept of fully renewable power systems, biomass is the ideal contender. The concept of CSP/biomass hybridisation relies on the ability of both systems to supply thermal energy in order to drive a power generation block. The advantages of this synergy go beyond dispatchability and renewable energy generation. They also include operation stability and flexibility, joint use of power plant equipment and associated cost reduction, as well as allowing CSP migration from desert areas to load centres [1,2,5–9]. Furthermore, during daylight periods, when electricity prices are usually higher, solar radiation is abundant and the system can run with larger solar shares resulting in a reduction of the levelised cost of energy. Despite the undeniable advantages of such systems, hybridisation still presents ambitious challenges, in order to guarantee that the system operates as a whole, and power reliability and stability are achieved. In this paper, a simulation model and results for a CSP/biomass hybrid mini power plant are presented. Annual simulations were carried out for solar-only and hybrid modes. Distinct operation ranges and boiler sizes were analysed. Simulation results are presented, such as: solar field annual generated heat and efficiency, boiler efficiency and biogas consumption, annual generated electrical energy and ORC efficiency, dumped heat, solar and biomass shares, and system global efficiency. Hourly results are presented for standard days, with and without hybridisation, showing the advantages of hybridisation. The solar field (SF) is constituted by parabolic trough collectors (PTC) with a net aperture area of 984 m2. As backup energy, a biogas boiler is used, fed through digestion of organic food waste. The nominal Organic Rankine Cycle (ORC) electrical output is 60 kW. The system was designed and a prototype will be installed in Tunis, in the framework of the REELCOOP project, co-funded by the EU [10]. 2. The hybrid mini power plant The hybrid mini power plant (Fig. 1) has a nominal electrical output of 60 kW, and relies on a regenerative ORC as power generation system, developed by Zuccato Energia. The turbine/generator block was adapted to assure operation at partial load, in order to compensate solar energy fluctuations, with a nominal gross efficiency reaching 13.3%. The ORC will be driven by saturated steam at 170 °C, which allows the power circuit to also operate
with available waste heat. Thermal generation will be achieved either from solar energy or biomass, or from the combination of both. The solar field relies on parabolic trough collector technology, and is constituted by 3 parallel loops of 4 PTMx/hp-36 collectors developed by Soltigua, with a net collecting surface of 984 m2. Direct Steam Generation (DSG) will be achieved in the solar field, and the recirculation concept was adopted. The solar field will be supplied with subcooled water and partial evaporation will take place in the solar collectors. The water/steam mixture is then separated in a steam drum, and therefore only saturated steam leaves the solar field. The leftover water is then recirculated. Complete evaporation enhances control complexity, implying unnecessary risks over solar the collector absorbers [11]. Auxiliary energy will be provided by a biogas steam boiler. The biogas will be produced by anaerobic digestion of canteen organic waste remains, showing a potential solution for the problem that waste disposal represents [12]. The system layout allows either hybrid or individual operation with each thermal source (solar-only or biogas-only). In order to reduce thermal energy waste as well as biogas consumption and to compensate short transients from solar power, a storage tank was foreseen in the project. Since the storage tank will be charged with saturated steam from the solar field, an isothermal latent thermal energy storage concept has been adopted. Whilst typical TES systems concern sensible heat storage with temperature change, the latent heat solution uses Phase Change Materials (PCM). 3. Simulation model The developed simulation model encompasses two stages. First, the solar field, the water/steam cycle and the boiler were analysed using a commercial software: EBSILONÒ Professional. This stage includes the simulation of the thermal generation system, for different operation profiles. The second stage concerns the power block circuit analysis that was carried out using EES software. The model required that relevant properties of the working fluid (Solvay SES36) were introduced, as they were not available either in the EES database nor in EBSILON. 3.1. Thermal generation system model EBSILONÒ Professional is a commercial simulation software designed for power plant and thermodynamic process analysis. The software includes a wide range of component libraries, for which steady state thermodynamic balances are simulated, under design and off-design conditions. The off-design calculations are carried out considering a linear relation between the off-design and design mass flow rate or/and thermodynamic properties (e.g. pressure, temperature, enthalpy). Simulations were carried out with an iteration precision of 107. In addition to the general EBSILON libraries, the EbsSolar, time-series, and EbsScript modules were used. The EbsSolar module consists of a component library (e.g. line focusing collectors, sun) specifically designed for solar thermal power plant simulations. To describe the solar power plant dynamic behaviour, the simulations were carried out, on an hour-by-hour basis, by using a time-series function. Furthermore, the detailed dynamic behaviour of the solar collectors and headers was considered. In real operation, the power plant is controlled depending on numerous external and internal variables. In order to describe this behaviour at different operation stages, a Pascal based code was developed using an EbsScript tool.
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
3
Fig. 1. Hybrid mini power plant layout [11].
Fig. 2. Thermal generation system simulation scheme.
The simplified system simulation scheme is presented in Fig. 2. The solar field model includes the sun, parabolic trough collectors, distributing and collecting headers and recirculating and feed water pumps. The biogas boiler was modelled throughout a heat injection component.
The model requires as input, metrological data (i.e. direct normal irradiation, ambient temperature). The data were obtained using Meteonorm software for the prototype site location (Tunis, Tunisia), on an hourly basis for a typical metrological year.
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
4
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
The Sun acts as interface between the meteorological data and incident radiation on the collectors. Sun angles, i.e. height (hs) and azimuth (cs) are calculated according to DIN-5034 standard, using as input the local time, latitude and longitude. In view of collector layout similarity, i.e. north-south orientation and non-slope, the normal irradiation incident angle (hi) and the collector tracking angle (q) can be determined [13],
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi cosðhi Þ ¼ 1 cos2 ðhs Þ cos2 ðcs Þ
ð1Þ
tan q ¼ sinðcs Þ= tanðhs Þ
ð2Þ
Therefore, the available beam irradiation at the collector surface (Ib), is a function of the incident angle (hi) and the direct normal irradiation (DNI),
Ib ¼ DNI cos hi
ð3Þ
For modelling the parabolic trough collectors, the line focusing solar collector component is used. The available energy within each collector (Qsolar) is defined as a function of the incident beam irradiation (Ib) at the net collector aperture area (Anet), considering the nominal collector optical efficiency (gopt), incident angle modifier (IAM), collector focus state (f), shading losses (gshad), and reflecting mirror cleanliness (gclean) [14],
Q solar ¼ Ib Anet gopt IAM f gshad gclean
ð4Þ
Manufacturer information concerning the nominal optical efficiency and incident angle modifier (IAM) were used for Soltigua’s collectors [15]. The IAM correction is a function of the incident angle, determined by a second order interpolation of table based values. When the sun height is low (e.g. sun appearing in the horizon), parallel collector row shading is expected to occur. This effect is determined by the collector row distance (lrow) and aperture width (Awidth) as well as the tracking angle (q) [14],
gshad ¼ 1 min ½1; max ð0; 1 lrow cos q=Awidth Þ
ð5Þ
A constant yearly mirror cleanliness factor (gclean) of 0.97 was considered [16]. The effective heat received by the heat transfer fluid (Qeff) is defined as the energy balance between the available solar energy (Qsolar) and thermal losses (Qloss). The heat losses at the collectors result from the temperature difference between the heat transfer fluid and the ambient air. A linear heat loss coefficient was considered according to manufacturer’s data [15]. The direct steam generation concept implies water evaporation within the collectors. Therefore, pressure losses (DPcoll) were calculated considering a two-phase flow model. The model entails the calculation of pressure losses considering a smooth pipe (DPcoll,S) and rough pipe (DPcoll,R). Afterwards, the worst case is considered:
DPcoll ¼ maxðDPcoll;S ; DPcoll;R Þ
ð6Þ
The smooth pipe calculations rely on the Friedel two-phase flow pressure drop model [17]. The model consists in the calculation of the pressure drop in the pipe as if it was a single-phase flow, both liquid and vapour. Subsequently, a correction factor is used, calculated from the Weber and Froude number for the liquid-phase flow as well as the steam quality within the pipe. On the other hand, the aforementioned model discards the pipe roughness. Thus, a correlation between single-phase and two-phase flow pressure drop in rough pipes is used. Initially, the pressure drop is calculated considering single-phase flow, for _ liquid (DPcoll,R,L) and gas (DPcoll,R,G), using the mass flow rate (m), pipe internal diameter (Dint), fluid density (qfluid) and a friction factor (f) [18],
DPcoll;RL
or G
¼ fL
or G
_2 m 2Dint qfluid
ð7Þ L or G
The friction factor (f) is a function of the pipe internal diameter (Dint), pipe roughness (e) as well as the Reynolds number (Re). In order to calculate f directly, the Swamee–Jain equation [19] is used,
fL
or G
2 e 5:74 ¼ 0:25 log10 þ 0:9 3:7Dint Re
ð8Þ
Thereafter the pressure drop in the absorber tube is calculated as a function of the steam quality (X),
DPcoll;R ¼ DPcoll;RL þ X ðDP coll;RG DPcoll;RL Þ
ð9Þ
In each case, the model uses the thermodynamics properties (e.g. enthalpy, pressure, steam quality) at the collector centre. For simulation purposes, a one-loop simplification was considered, in order to reduce the computational effort. The other two loops were accounted through the use of distributing and collecting headers. These components allow considering the fluid stream distribution and collection, without neglecting mass flow, heat and momentum balances. Whilst mass flow rate within the loops is considered identical, enthalpy and pressure are different due to heat and pressure losses along the header. The header pressure loss model is identical to the one for solar collectors, with a single-phase flow consideration in the distributing header and a two-phase flow in the collecting header. As inputs, the header length (60 m) and a constant diameter were considered, divided into three identical sections. The internal diameters of the distributing and collecting headers, were defined as 0.0418 and 0.053 m, respectively. Heat losses (Qloss,H) within the header are calculated by the radial heat conduction in the insulting material [20],
Q loss;H ¼ 2pkins
T T amb fluid ln Dext;ins =Dint;ins
ð10Þ
Calculations are carried out for each section, with fluid temperature (Tfluid) determined at the section centre. The insulation thermal conductivity (kins) was defined as 0.1 W/mK and the ratio between the outer (Dext,ins) and inner (Dext,ins) diameters as 2.24. Heat and pressure losses in the pipe connections within the loop are considered as well. Pressure drop is calculated by the application of Bernoulli’s theorem for a specific pipe geometry. Heat losses are calculated using Eq. (10). The transient behaviour of the solar field is modelled through the use of an indirect storage (IS) component. The implementation of an IS component was restricted to each of the solar collectors and both headers, where the highest temperature gradients are expected to occur. The IS model calculates the transient heat exchange in a pipe between the working fluid (i.e. water/steam) and either the collector absorber tube or piping. The Fourier heat transfer differential equation [20],
@T @ 2 T pipe @ 2 T pipe qcp pipe ¼ k þ @t @x2 @y2
!
ð11Þ
is discretised in a two-dimensional space domain using a finite volume method, and in time by an iterative Crank-Nicholson method [21], The convective heat flux (Q_ pipe;fluid ) between the pipe walls and the heat transfer fluid is calculated using the Newton’s cooling law [20], for each control volume area (A) and assuming a constant heat transfer coefficient along the pipe,
Q_ pipe;fluid ¼ aAðT pipe;wall T fluid Þ
ð12Þ
The dynamic behaviour of the fluid temperature along the pipe, is modelled by the energy and mass balance:
_ in hout Þ mfluid cp;fluid mðh
@T fluid ¼ Q_ pipe;fluid @t
ð13Þ
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
A higher number of elements in the mesh, significantly improves simulation accuracy, however at the expense of increased computational effort. Based on a sensitivity analysis and considering that the temperature gradient is much higher in the axial direction, the control volume was divided into 60 elements (i.e. 30 in the axial direction and 2 in the radial direction). The time-step (Dt) was defined considering the Courant–Friedrichs–Le wy stability condition [22], for each element size (Dx), using the fluid velocity (Vfluid) at the pipe centre,
V fluid Dt=Dx < 1
ð14Þ
The nominal convective heat transfer coefficient (a) used in Eq. (12), was calculated for each collector and header,
a ¼ kNuD =Dint
ð15Þ
The Nusselt number (NuD) was estimated by the Dittus-Boelter equation, using the Reynolds (Re) and Prandtl (Pr) numbers,
NuD ¼ 0:023Re0:8 Pr 0:4
ð16Þ
3.1.1. Hybridisation Hybridisation was modelled through a heat injection component, representative of the boiler. This component acts as an ideal heat exchanger, promoting the interface between the boiler output and the mass flow rate of water/steam. In order to control the boiler output, a code was created using EbScript to impose operating limits according to the manufacturer data, as well as to calculate the required heat flow. The boiler output is controlled by the water/steam mass flow rate, acting as an auxiliary heater of the solar field. For simulation purposes, the Viessmann VITOMAX 200-HS boiler model with economiser was used. Two different boiler sizes were the object of analysis, with nominal heat outputs of 380 kWth (0.5 ton/h of saturated steam) and 530 kWth (0.7 ton/h of saturated steam), and both with a minimum thermal output of 100 kWth. In order to account for the transient behaviour of the boiler an IS component was used. As inputs, the water volume of the boiler was used and the mass of steel estimated, considering that the boiler takes half an hour from cold start to design conditions. For estimating biogas consumption, an additional computer model was created using Ebsilon, mainly constituted by a combustion chamber and two heat exchangers (i.e. steam evaporator and economiser) (see Fig. 3). The model uses the water/steam mass flow rate, as well as the inlet and outlet enthalpies to calculate the required boiler output. These inputs are obtained from the thermal generation system model results. The boiler efficiency, combustion heat output and flue gas temperature are determined as a function of the ratio between the boiler output and the rated output, considering the
5
manufacturer data. In the combustion chamber, the combustion of the mix of air and biogas is modelled, considering a 3% oxygen excess in the flue gas. As output it retrieves the biogas mass flow rate. Biogas consumption was estimated considering a yearly constant low heating value (19.27 MJ/m3). A study was conducted at ENIT, for estimating biogas production with the local canteen residues. The estimated value is about 60 m3/day. 3.2. Power block The regenerative Organic Rankine Cycle was simulated using EES, since property data of the organic fluid (Solvay SES36) were not available neither in the EES nor in the Ebsilon databases. The data values were introduced into tables from which the thermodynamic properties (e.g. enthalpy, entropy) are calculated by linear interpolation. It was assumed that the ORC operates at steady-state, with thermal inertia of the power block neglected. Furthermore, the evaporation and condensation temperatures are considered constant, as well as the pump efficiency and the temperature difference in the regenerator. As input the code requires the saturated steam mass flow rate obtained from the thermal generation system simulation results. The steam mass flow rate is used to calculate the turbine isentropic efficiency, by means of a second order equation based on manufacturer data. Within the ORC model, calculations for individual components (i.e. turbine, pump and heat exchangers) are achieved by applying the first law of thermodynamics. Irreversibilities are considered at the turbine and pump. As main output results the code provides the gross and net electrical power, organic fluid mass flow rate, as well as the parasitic consumption and the condenser thermal requirements. 3.3. Operation modes and control In order to optimise the system operation and to enhance similarity with real operation, the thermal generation system model is controlled and also distinct operation profiles are considered. EBSILON contains a wide-range of components suitable for control purposes, such as property value transmitters, calculators and simple linear controllers. Nevertheless, more complex control tasks, such as interchange between operation profiles, require the development and implementation of a code. Therefore, a computational code was developed, using a Pascal based language and implemented in the numerical model using the EbScript tool. 3.3.1. Solar-only operation mode The system steam production will be controlled by the mass flow balance at the steam drum. The same concept was used in the simulation model. During system operation the recirculating
Fig. 3. Boiler simulation scheme.
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
6
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
mass flow rate is kept constant at about 0.5 kg/s. In real conditions the feed water pump will operate when the water level in the steam drum drops below a predefined level. Since the simulations were carried out on an hourly basis, it was assumed that the feed water mass flow rate should balance the water removed from the recirculating pump, and so is intrinsically related to the steam quality after the collecting header. If the saturated steam mass flow rate that leaves the solar field exceeds the maximum requirements of the ORC, it is then separated using a simple splitter (see Fig. 2). In real operation, the solar collectors should change their state to partial defocused in order to control the steam production. In the simulation model, the excess of energy is accounted as dumped, providing an input for a PCM storage tank design. The last separation occurs before the ORC, through a water-steam drainer. This component is mostly used in the warm-up and cool-down profiles, to establish a more realistic thermodynamic balance. To obtain a more accurate approach in the simulation model, distinct operating profiles were created. A code was developed using EbScript to allow the automatic interchange between the profiles using dynamic variables, e.g. direct normal irradiation, hour of the day, mass flow rates, etc. The solar-only operating mode is constituted by four profiles: warm-up, operating, cool-down and stop. At the beginning of the day if DNI exceeds 200 W/m2 the collectors change their state to focus and the recirculating pump is activated. This represents the warm-up profile. The operation profile starts with solar field steam production, and operates as described before. At the end of the day, if DNI is less than 200 W/m2, the collectors change their state to defocus and water circulates until the system cools-down. At night, the system is off, with solar field collectors defocused and the pumps shut-down. Even during the night, the thermal inertia of the system as well as the heat losses to the ambient were considered. 3.3.2. Hybrid operation mode The hybrid operation analysis was carried out based on the assumptions of the system running 12 or 24 h daily, at ORC minimum and nominal power. Concerning the control, the hybrid mode comprehends four and two operating profiles for the 12 and 24 h regimes, respectively. The 12 h operating regime differs from the solar-only, on the warm-up and operating profile. The warm-up begins at 7:00 with the start of the boiler in order to warm-up the ORC and the feed water pipe of the solar field. If DNI exceeds 200 W/m2 the recirculating pump is activated and the solar collectors focused. During the operation regime (08:00–20:00) the hybrid mode is activated, with the boiler compensating the requirements (nominal or minimum) to drive the ORC turbine. This control is achieved by saturated steam flow rate balance. In preliminary simulation results it was noticed that during summer the system could start earlier (at 7:00), and the 12-h operation regime acted as a constraint to the solar field. To overcome this issue, in the beginning of the day if DNI is above 200 W/m2 the collectors are focused earlier. The 24-h regime just encompasses two operation modes, hybrid and boiler-only. During the daylight period the system operates in hybrid mode, and at night the collectors are defocused and the solar field is cooled down. After that the system relies solely on the boiler. For both cases the boiler minimum power of 100 kWth was considered, in order to assure electricity generation stability, during the predefined time operating range. Otherwise, the boiler would be submitted to consecutive start-ups and shutdowns, and shortages in the electrical generation would be expected, due to solar radiation transients.
4. Results If the system relies solely on solar energy the annual heat generated is about 663 MWhth. Almost one quarter of the heat is dumped, mostly related with energy dearth. The dumping rate results are divided in two items (excess and scarcity), representing the heat dumped due to the excess of energy or due to insufficient energy to drive the ORC turbine, respectively. The annual power generated is 61 MWhel with an ORC average annual efficiency of 9.2% and an annual running time of 1420 h. The system efficiency is 3.4% hindered by the excessive dumping rates (see Table 1). Hybridisation improved the solar field output by 3% (Table 2). This outcome is related with the system start-up, since the SF feed water is already warmer, and consequently less solar energy is required to achieve steam generation. Furthermore, this improvement is extended to the solar field annual efficiency. The second improvement of hybridisation was the extinction of dumping rates associated with scarcity of energy. The scarcity of energy was surpassed with hybridisation, with the fulfilment of the ORC minimum thermal power requirements. Moreover, ORC operation near nominal conditions enhanced the ORC efficiency. On the other hand, the excess of energy increased. This fact is related with minimum operating conditions of the boiler (100 kWth), leading to energy waste predominantly in the summer months, when solar radiation is highly available. The excess of energy can be reduced by implementing a storage tank in the system. The benefits extend beyond the ability to store the excess of solar field thermal energy. If the storage can provide more than 30 min of thermal energy requirements to drive the ORC turbine, the need of having the boiler in permanent operation is eliminated. In other words, it can act as system buffer in order to compensate thermal output fluctuations from the solar field and boiler, reducing the amount of wasted biogas. Despite the discontinuous operation of the boiler, due to the solar irradiance transients, the average biogas boiler efficiency is still high (about 93%) for all cases. As expected, electric generation is significantly increased with hybridisation. If the system depends only on solar energy, generation is confined to sunnier months. In spite of the favourable Tunisian climate, solar radiation monthly values show a noteworthy discrepancy between summer and winter. To enable electrical generation in low radiation months, the solar multiple should be increased, leading to costly and oversized systems. On the other hand, even system hybridisation for 12 h/day at ORC minimum power showed an annual generation of 167 MWhel, with a solar share of 44%. System annual efficiency experienced a huge boost with hybridisation, from 3.4% to almost 10%, due to the high efficiency boiler, along with improved efficiencies of the SF and ORC
Table 1 Solar-only simulation annual results. Direct Normal Irradiance – DNI Annual Heat Generated S.F. Specific Thermal Field Output Mean Annual SF Efficiency Dumping Rate – Excess Dumping Rate – Scarcity Annual Useful Heat – ORC Annual Power Generated Mean Annual ORC Efficiency ORC – Number of hours running Annual Dissipated Heat – Condenser Maximum Heat Dissipated – Condenser Mean Annual System Efficiency
1799.4 663 674 37.4% 7.9% 16.3% 503 61 9.2% 1420 437 387 3.4%
kWh/(m2 a) MWhth kWhth/m2 % % % MWhth MWhel % h MWhth kWhth %
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
7
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx Table 2 Hybrid simulation annual results. Hybrid – 24 h operation
Direct Normal Irradiance – DNI Annual Heat Generated S.F. Specific Thermal Field Output Mean Annual SF Efficiency Annual Heat Generated – Boiler Annual Combustion Heat – Boiler Mean Annual Boiler Efficiency Annual Biogas Consumption Average Biogas Consumption Solar Share Dumping Rate – Excess Dumping Rate – Scarcity Annual Useful Heat – ORC Annual Power Generated Mean Annual ORC Efficiency ORC – Number of hours running Annual Dissipated Heat – Condenser Maximum Heat Dissipated – Condenser Mean Annual System Efficiency
Hybrid – 12 h operation
530 kWth
380 kWth
ORC minimum
530 kWth
380 kWth
ORC minimum
Boiler
Boiler
Power
Boiler
Boiler
Power
1799.4 683 694 38.6% 3359 3619 92.8% 655 1795 17% 2.5% 0.0% 3941 515 12.7% 8760 3389 387 9.6%
1799.4 683 694 38.6% 2546 2746 92.7% 497 1362 21% 3.1% 0.0% 3128 380 11.8% 8760 2720 387 8.4%
1799.4 683 694 38.6% 1788 1918 93.2% 347 951 28% 4.1% 0.0% 2370 261 10.6% 8760 2089 387 7.1%
1799.4 683 694 38.6% 1425 1534 92.9% 278 761 32% 6.5% 0.5% 1971 257 12.2% 4380 1695 387 7.8%
1799.4 683 694 38.6% 1155 1244 92.8% 225 617 37% 7.5% 0.5% 1700 213 11.6% 4380 1472 387 7.1%
1799.4 683 694 38.6% 865 928 93.2% 168 460 44% 8.9% 0.0% 1410 167 10.8% 4380 1231 387 6.2%
kWh/(m2 a) MWhth kWhth/m2 % MWhth MWhth % dam3 m3/day % % % MWhth MWhel % h MWhth kWhth %
should be used only during start-up and shut-down, reducing the amount of biogas consumption and heat dumped. The boiler starts to operate at 07:00, at minimum power in order to warm-up the SF and ORC. From 8:00 to 20:00 nominal electricity generation is achieved mostly from solar energy. Either Summer or Winter solstice results, showed that electricity stabilisation can be achieved during the daylight period through hybridisation, moreover with a significant solar share in the summer (75%) and less relevant (13%) in the winter. This synergy is noteworthy, since during the daylight period electricity prices are usually higher, related to peaks in network consumption. The estimated biogas production of 60 m3/day is far below the consumption results. On an annual basis the excess of energy related with the boiler minimum operating conditions represents 738 h of operation and 14 dam3 per year. If we consider as an example the case of 12 h of operation at minimum power, this represents 9% of the annual biogas consumption and 16% of the running hours. This denotes a minor contribution to CSP/biomass hybridisation. Despite hybridisation allowing to relocate a power plant near urban centres where organic wastes are more abundant,
(Fig. 4). In the REELCOOP framework a 10% system efficiency was proposed as target, which is nearly attainable with hybridisation. One of the main advantages of hybridisation is the stability of the system, in order to promote dispatchability. This can be observed in the simulation for the 21st of December (Fig. 5), where the generator is operating 24 h at nominal power with the 530 kWth boiler. On the 21st of December, the boiler is supplying 450 kWhth until 9:00, fulfilling the ORC requirements during night operation. At this time the DNI is above 200 W/m2, and the collectors change their sate to focus. Steam production from the solar field starts at 10:00, with a thermal output of about 50 kWth. When compared with solar-only, the steam generation starts one hour later and with half of the hybrid production. During the day the solar field is unable to supply the minimum conditions to drive the ORC turbine. Nevertheless, it contributes to a reduction in the amount of required boiler energy and biogas consumption. On the Summer solstice (21st June), the generation stability was attained even with the smaller boiler (Fig. 6). However, the heat dumped due to excess of energy increased. In such days, the boiler Annual Power Generated [MWhel] Proposed Target Efficiency [%]
Mean Annual System Efficiency [%] 10%
10% 8.4% 7.8% 7.1%
7.1% 6.2%
3.4%
9.6% 515
380 257
261
530 kWth Boiler
ORC minimum Power
213 167
61
Solar Only
ORC minimum Power
380 kWth Boiler
12 hours
Hybrid - NP Boiler 1
530 kWth Boiler
24 hours
Fig. 4. Annual power generated and mean annual efficiency.
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
8
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
Solar Field Thermal Output -Hybrid
NET Power ORC
Boiler Thermal Output
Solar Field Thermal Output - Solar Only
500
900
450
800
400
700
350
600
300
500
250 400
200
DNI [W/m2]
Thermal and Electric Output [kW]
DNI
300
150 100
200
50
100
0
0 0
2
4
6
8
10
12
14
16
18
20
22
Hours
Solar Field Thermal Output
Boiler Thermal Output
NET Power ORC
DNI
500
900
450
800
400
700
350
600
300
500
250 400
200
DNI [W/m2]
Thermal and Electric Output [kW]
Fig. 5. Winter solstice 24-h hybrid operation with a 530 kWth boiler.
300
150 100
200
50
100
0
0 0
2
4
6
8
10
12
14
16
18
20
22
Hours Fig. 6. Summer solstice 12-h hybrid operation with a 380 kWth boiler.
500
Mean Daily Heat Dumped - Excess [kWhth]
Mean Daily Heat Dumped - Scarcity [kWhth] 435
450 400
371 325
350
332
325
300
322
273 239
250
208
184
200
224
231 202
150 76
76
48 0
130
121
100 50
382
370 342
36
17 January
10 February
March
April
May
June
July
August September October November December
Fig. 7. Heat dumped in the solar-only operation mode.
the absence of a well-established biomass market represents a drawback to the dissemination of these systems [1]. As aforementioned, either solar-only or hybrid operation results showed excess of thermal energy, mostly in the summer months
when solar radiation is abundant. The energy waste can be overcome with a storage tank. To define the ideal storage capacity the daily average values of the heat dumped due to energy excess were analysed for the solar-only mode (Fig. 7). The analysis was
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180
J. Soares, A.C. Oliveira / Applied Thermal Engineering xxx (2016) xxx–xxx
9
not extended to the hybrid mode, since it enhances the dumped heat and consequently the storage capacity. Furthermore, this overestimation can easily be eliminated with a small storage capacity. The maximum value for the average daily dumped heat is 332 kWhth (June) and the minimum is 10 kWhth (December). It is worth to note that the maximum daily value of the excess heat does not allow to drive the ORC at nominal power for one hour.
grant agreement no. 608466 (www.reelcoop.com). All other partners involved in the development of Prototype 3 are greatly acknowledged: German Aerospace Centre (DLR, Germany), CIEMAT (Spain), ENIT (Tunisia), Soltigua (Italy), Zuccato Energia (Italy), AES (Tunisia).
5. Conclusion
[1] A. Colmenar-Santos, J.-L. Bonilla-Gómez, D. Borge-Diez, M. Castro-Gil, Hybridization of concentrated solar power plants with biogas production systems as an alternative to premiums: the case of Spain, Renew. Sustain. Energy Rev. 47 (2015) 186–197. [2] H.G. Jin, H. Hong, 12 - Hybridization of concentrating solar power (CSP) with fossil fuel power plants, in: K. Lovegrove, W. Stein (Eds.), Concentrating Solar Power Technology, Woodhead Publishing, 2012, pp. 395–420. [3] B. Coelho, S. Varga, A. Oliveira, A. Mendes, Optimization of an atmospheric air volumetric central receiver system: impact of solar multiple, storage capacity and control strategy, Renewable Energy 63 (2014) 392–401. [4] G.K. Singh, Solar power generation by PV (photovoltaic) technology: a review, Energy 53 (2013) 1–13. [5] A.A.A. Cot, J. Vall-Llovera, J. Aguiló, J.M. Arqué, Termosolar Borges: A Thermosolar Hybrid Plant with Biomass, in: Third International Symposium on Energy from Biomass and Waste, CISA, Environmental Sanitary Engineering Centre, Italy, Venice, Italy, 2010. [6] Á. Moreno-Pérez, P. Castellote-Olmo, Solar Parabolic Trough – Biomass Hybrid Plants: Features and Drawbacks, in: International Conference on Concentrating Solar Power and Chemical Energy Systems, SolarPACES 2010, Perpignan, France, 2010. [7] B. Coelho, P. Schwarzbözl, A. Oliveira, A. Mendes, Biomass and central receiver system (CRS) hybridization: volumetric air CRS and integration of a biomass waste direct burning boiler on steam cycle, Sol. Energy 86 (2012) 2912–2922. [8] J.H. Peterseim, S. White, A. Tadros, U. Hellwig, Concentrated solar power hybrid plants, which technologies are best suited for hybridisation?, Renewable Energy 57 (2013) 520–532 [9] J.H. Peterseim, S. White, A. Tadros, U. Hellwig, Concentrating solar power hybrid plants – enabling cost effective synergies, Renewable Energy 67 (2014) 178–185. [10] REELCOOP, REELCOOP project, in, 2015. [11] D. Krüger, A. Kenissi, S. Dieckmann, C. Bouden, A. Baba, A. Oliveira, J. Soares, E. R. Bravo, R.B. Cheikh, F. Orioli, D. Gasperini, K. Hennecke, H. Schenk, Pre-design of a mini CSP plant, Energy Procedia 69 (2015) 1613–1622. [12] A.C. Oliveira, B. Coelho, REELCOOP Project: Developing Renewable Energy Technologies for Electricity Generation, in: 12th International Conference on Sustainable Energy Technologies (SET2013), Hong Kong, 2013. [13] W.B. Stine, R.W. Harrigan, Solar Energy Fundamentals and Design: With Computer Applications, John Wiley & Sons, Incorporated, 1985. [14] STEAG Energy Services GmbH, Line focusing Solar Collector, in: Solar Library EbsSolar, 2015. [15] Soltigua, PTMx/hp Parabolic Trough Collector – Technical Data Sheet, in, 2013. [16] I. Llorente García, J.L. Álvarez, D. Blanco, Performance model for parabolic trough solar thermal power plants with thermal storage: comparison to operating plant data, Sol. Energy 85 (2011) 2443–2460. [17] H. Schmidt, A. Wellenhofer, S. Muschelknautz, J. Schmidt, F. Schmidt, D. Mewes, A. Mersmann, J. Stichlmair, L2 Two-Phase Gas-Liquid Flow, in: VDI Heat Atlas, Springer, Berlin, Heidelberg, 2010, pp. 1117–1180. [18] B.R. Munson, A.P. Rothmayer, T.H. Okiishi, Pre-design of a Mini CSP Plant, seventh ed., John Wiley & Sons Incorporated, 2012. [19] P.K. Swamee, A.K. Jain, Explicit equations for pipe flow problems, J. Hydraul. Eng. 102 (1976) 657–664. [20] F.P. Incropera, T.L. Bergman, D.P. DeWitt, A.S. Lavine, Fundamentals of Heat and Mass Transfer, Wiley, 2013. [21] R. Pawellek, S. Pulyaevrank, T. Hirsch, Transient simulation of a parabolic trough collector in EBSILON, in: International Conference on Concentrating Solar Power and Chemical Energy Systems, SolarPACES 2012, Marrakesh, Morocco, 2012. [22] R. Courant, K. Friedrichs, H. Lewy, Über die partiellen Differenzengleichungen der mathematischen Physik, Math. Ann. 100 (1928) 32–74.
In this paper, a simulation model and results of a hybrid CSP/biomass mini power plant were presented. The simulations were carried out on an hour-by-hour basis for a typical meteorological year, considering different scenarios regarding operating regimes and boiler sizes. If power generation is exclusively dependent on the solar field, electrical generation is mostly confined to sunnier months and negligible in the winter. The annual heat dumping rate ascends to almost one quarter, mostly related with the inability of the solar field to supply the minimum thermal energy to drive the ORC turbine. System hybridisation proved to stabilise the system regarding electrical power generation during the whole year. Additionally, the downside of the dumped heat, due to scarcity of thermal energy, was surpassed. Hybridisation improvements where extended as well to the SF and ORC efficiencies. The SF efficiency increased 3% since the system is already warmer in the morning, and solar radiation is used exclusively for steam generation. The ORC efficiency increase is in the range of 15 to 38%, and was achieved by a stable operation near turbine design conditions. On the other hand, there was an increase in the system energy excess due to the downsides of boiler start-up time (about 30 min) and minimum operational heat input (100 kWth). The simulated operation profiles were created on the basis of energy dispatchability and demand response ability. Therewith, boiler operation was extended to the system operation range to compensate possible short transients from solar power. This issue can be overcome with the implementation of a storage tank. From simulation results, the maximum average daily dumped heat by excess ascends to 332 kWhth. This value is quite small when compared with the system size. For example, it does not allow to run the system for one hour at nominal power. This proves that the solar field design is appropriate and hybridisation significantly reduces (although not totally eliminating) the need of storage. The improvements in the SF and ORC efficiencies, along with extended operation ranges and a highly efficient boiler, lead to a huge boost on system efficiency, which increases from 3.4 to 9.6%. As future work, system simulations will include the implementation of a PCM storage tank. More operation strategies, biomass resources, different climatic regions and a cost analysis will be addressed. A prototype of the system was developed and will be installed in Tunisia, for testing during 2016/17, allowing a validation of the simulation results.
References
Acknowledgements The REELCOOP project receives funding from the European Union Seventh Framework Programme (FP7/2007-2013), under
Please cite this article in press as: J. Soares, A.C. Oliveira, Numerical simulation of a hybrid concentrated solar power/biomass mini power plant, Appl. Therm. Eng. (2016), http://dx.doi.org/10.1016/j.applthermaleng.2016.06.180