Optimal design of a concentrated solar power plant with a thermal energy storage

23 de diciembre de 2023

In this work, a concentrated solar power (CSP) plant with a thermal energy storage system to produce 120 megawatts of electrical energy was designed using Thermoflex. As non-renewable resources are limited, to move towards sustainable development, the importance of developing alternative technologies such as optimized solar power generation systems becomes considerable. To eliminate the gap between electricity demand and power supply in solar power plants, it is necessary to use optimized thermal energy storage systems. So, in this study the Nelder-Mead algorithm as a novel optimization method was coupled with Thermoflex in the Excel environment to obtain an optimized techno-economic design. Finding the minimum of the Rosenbrock benchmark function confirmed that the simplex algorithm can find optimal solutions in a short time for complicated optimization problems. For plant optimization, twelve parameters that have the greatest effect on the objective function were selected as decision variables. The optimization results after 77 iterations demonstrated that in the optimal case, the objective function decreases from 4.48 to 4.38. This was equivalent to a 2.98 MWe power increase, while the cost of CSP components remained at an acceptable level. The total heat transfer for the solar boiler and feed water heater train sections before and after optimization is 391.816 MWth and 395.033 MWth, respectively.

Introduction

The non-renewable energy resources in the world are severely restricted, and the energy crisis has become the most crucial challenge for humanity in the current age [1], [2], [3]. The use of alternative energies, such as renewable energies, can overcome this crisis [4], [5]. In addition, the destructive effects of non-renewable energies, especially fossil fuels, on the environment have increased the importance of developing energy technologies [6].

It seems that solar energy can be one of the best alternatives to non-renewable sources due to its unlimited and non-polluting nature [7], [8]. The amount of energy from sunlight reaching the earth in one hour is equivalent to the energy used by people in one year [9]. This form of energy can be converted into electrical energy using photovoltaic (PV) and concentrated solar power (CSP) [10]. PV directly converts sunlight into electricity, while CSP harnesses its heat and converts it into electrical energy under a thermodynamic cycle. The CSP includes four technologies. The type of mirror used in any technology is different. Parabolic Trough technology (PT) is the first one. In this type, parallel rows of reflectors or mirrors are assembled, and then stainless-steel absorber tubes are used as heat collectors. The second concentrated solar power technology is the solar tower (ST). Solar tower power plants and PT power plants have almost identical structures. But ST technology uses many heliostats. ST technology creates the highest degree of temperature among the other CSP technologies. The most important advantage of this technology is the possibility of hybrid operation. The possibility of reaching high temperatures and hybrid operation are the main benefits. The third is the Fresnel mirror system. In this technology, mirrors focus on a focal point. Water flow in pipes runs directly through the focal point. The fourth technology is Parabolic Dishes (PD). The PD concentrates the sun’s rays at a focal point in the center of the dish [11]. The worldwide capacity of solar energy installations by these methods from 2010 to 2020 increased from 41,600 to 716,153 MWp [12]. Solar thermal power plants use direct solar radiation (DNI). This part of the sun’s radiation is deflected by clouds, aerosols, or dust. Therefore, these plants should be constructed in areas with sufficient solar radiation [11]. Suitable sites for the construction of solar power plants have solar radiation of 2000 kWh/m2y annually. One of the upcoming challenges in using solar energy is its interruption at night or on days without sun [13]. Currently, the available solution to overcome this problem is to use backup systems or thermal energy storage technologies. The use of backup systems such as fossil systems, while moving us away from the goal of producing power without pollution, increases the costs to a great extent. Therefore, the use of energy storage systems is more practical and economical [14]. Round-the-clock power generation to eliminate the gap between supply and demand of electricity using thermal energy storage (TES) is a technology that has made CSP superior to photovoltaic and other renewable sources. In general, TES is less expensive than electrical energy storage in batteries [15].

The results of research on 85 operated photovoltaic power plants in Central Europe show that their actual lifetime is about half that of the originally planned lifetime. This greatly affects their economy. Investigations also show that the huge initial costs of CSPs lead to a long capital recovery period, which may add to the investment risks. Meanwhile, it is still economically feasible due to the better-expected benefits [16], [17]. The cost-effectiveness of fully dispatchable solar energy with CSP technology depends on the optimal design of these systems [18]. Therefore, for an optimal design, cost and production efficiency should be considered simultaneously. For the stated reasons, a lot of research has been done in the field of harnessing this endless source of energy.

Solar energy research covers PV [19], [20], [21], CSP [22], [23], [24], and integrated systems [25], [26], [27]. In 2021, Bousselamti et al. optimized an integrated system of CSP and PV using the genetic algorithm with the objective of determining the minimum levelized cost of electricity (LCOE) and maximum capacity factor. The high energy dispatchability of a CSP system and the low cost of PV technology have been implemented in this integrated design [28]. Agyekum et al. evaluated the technical and economical proficiency of two practical CSP technologies: solar towers (ST) and parabolic troughs (PT) at two different places in Ghana (Navrongo and Tamale). The results showed that the TES period is an important factor in determining the LCOE for both technologies [29]. In addition to these, many studies have been conducted in the field of solar-gas hybrid systems, solar-wind systems, and other types of hybrid systems, all of which aim to produce power in optimal conditions without environmental pollutions [30], [31], [32], [33], [34]. In some of the hybrid systems, the environmental impact of the system has been investigated in terms of the reduction in CO2 emissions per year and the credits gain as a result of this reduction [35], [36]. Soltani et al investigated an energy storage system based on liquid air energy storage (LAES) and high-temperature concentrated solar power (CSP). Their study includes energy, exergy, economic, environmental, and sustainable analysis. In this regard, by using parametric analysis, the most important system performance parameters have been introduced, and the sources of energy destruction in different units have been optimized to prevent energy loss [37].

So far, many of the researchers designed CSP plants using various software. However, there is still a lack of designs to be closer to the real conditions in terms of economic, energetic, and operational aspects using power full software such as Thermoflow. Thermoflow’s world-renowned software is used by thousands of power plant engineers at hundreds of companies around the world. Whether they are designing new plants, repowering old facilities, or analyzing how to best operate their fleet of power plants, they use Thermoflow software to do the jobs. In addition, there is a lack of research on the multi-objective optimization of CSP plants. The present work aimed to design and optimize a large CSP plant with 120 MWe of net power. Among the four types of CSP technologies, the solar tower (ST) is most easily adapted to power generation cycles such as Rankine and Brayton. So, it has been used as a heat source for steam production. A TES tank containing molten salts is added to the first loop to reduce the gap between power demand and supply. To perform the optimization procedure, the CSP plant has been designed in Thermoflex. In this paper, a novel local search derivative-free algorithm, the Nelder-Mead, is used for CSP plant optimization. This optimization is affected by thermodynamic conditions and evaluation of several functions. A deeper look into the interdependencies of the solar field, storage, and power block shows that mutual sub-system influences can be very strong and indirect. For instance, when an optimized solar field operation temperature is determined, it is important to consider such interdependencies. Until now, iterative calculations using several programs were necessary to assess mutually influencing effects. Alternatively, the new integrated approach, as presented here (coupling the Thermoflow with the optimization algorithm in the Excel environment), can address this complexity through integral modeling and a powerful multi-objective optimization algorithm. This approach is applied to CSP systems.

Section snippets

Nelder–Mead (downhill simplex method) algorithm

The Nelder–Mead technique, or simplex search optimization method, was presented by John Nelder and Roger Mead in 1965 [38]. This numerical method minimizes or maximizes, a nonlinear unconstrained function f(x) without derivative information; the simplex has N-dimensional space and N + 1 vertices x(1), x(2), … x(n + 1), where f(1) < f(2) < … < f(n + 1). x(n + 1) and x(1) are initialized as the worst point and the best point, respectively. To minimize the objective function, the best and worst

Specification of the proposed plant

The proposed plant based on the heliostat field is located at latitude 33°.07′ N, longitude 5°.31′ E, and altitude 304.8 m. The total direct average irradiation is greater than 9.877 KWh/m2/ day at an ambient temperature of 29.4 °C. For this location, the June average daily radiation has been illustrated in Fig. 5. The thermal power output of the CSP plant is calculated according to equation (2) [42]:

Where DNI is the direct normal irradiation,

is the total heliostat

Optimization procedure

The most important challenge for the development of solar power plants is the component cost of the plant. In fact, one of the key factors that determines the development of CSP technology is its cost-effective and efficient design, together with thermal energy storage systems (TES) [44]. Recently, with the implementation of gradient metal foams into phase-change materials and Nano-materials for energy storage systems, advances in TES have been obtained [43], [45]. But to increase the

Results

The proposed CSP plant presented in Section 3 has been optimized by defining twelve decision variables and one fitness function (Equation (7). The constraints of the fitness function have been reported in Table 3. The process of reducing in the fitness value in successive iterations is plotted in Fig. 7. From the beginning of the optimization to the 77th iteration, the fitness function decreases with a significant slope; after that, it remains approximately constant. So, to save time, the

Conclusion

This study has been conducted with the aim of optimizing a CSP plan along with a TES system using the Downhill Simplex algorithm. In this study, solar energy has been chosen as one of the best alternatives to non-renewable sources due to its unlimited and non-polluting nature. Since among the four types of CSP technologies, ST is easily compatible with power generation cycles such as Rankine and Brayton, it is chosen as the heat source for steam generation. The simulation of the power plant was

CRediT authorship contribution statement

A. Naserbegi: Data curation, Formal analysis, Software, Writing – original draft. M. Aghaie: . Z. Nourani: Data curation, Formal analysis, Software, Writing – original draft.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (48)

https://www.sciencedirect.com/science/article/abs/pii/S1359431123022937

Recibe nuestras noticias por email

Recibe todas las novedades publicadas en nuestro portal de noticias.