Development of an integrated tool to design, estimate cost and calculate annual performances of a solar power tower

(1)

Development of an integrated tool to design, estimate cost and calculate annual performances of a solar power

tower

Emil Blampain

Master of Science Thesis

KTH Industrial Engineering and Management Energy Technology EGI_2018_0101-MSC EKV1222

Division of Heat and Power Technology SE-100 44 STOCKHOLM

(2)

(3)

Examensarbete EGI_2018_0101-MSC EKV1222 Utveckling av ett verktyg som kan utforma,

beräkna kostnaden och beräckna årliga avkastningar på ett smält salt soltorn

Emil Blampain

Godkänt Examinator

Björn Laumert

Handledare

Björn Laumert

Uppdragsgivare Kontaktperson

Sammanfattning

Denna uppsats bestod i att genomföra ett verktyg som kan utforma, beräkna kostnaden och beräkna årliga avkastningar på ett smält salt soltorn. Ett sådant verktyg gjordes för ett företag inom soltornsteknik, upphandling och konstruktion (SUK) som vill föreslå konkurrenskraftiga anläggningskonfigurationer som presenterar en bra avvägning mellan kostnad och intäkter.

Företaget, samtidigt som det övervakar SUK för en hel kraftverk, levererar det också vissa komponenter i den smälta saltcykeln. Verktyget modellerar ett storskaligt soltorn med ett värmeenergilagringssystem på EBSILON®Professional 12.04, en termodynamisk programvara.

När en simulering startas, ritar verktyget komponenterna i den smälta saltcykeln (designfas) enligt användarens inmatningar, de andra komponenterna är baserade på ett referensprojekt. Beroende på komponenternas storlek bestäms den totala kostnaden och intäkterna över ett verksamhetsår beräknas (årlig prestation). När flera simuleringar görs med olika konfigurationer kan företaget bedöma sin ekonomiska lönsamhet genom att jämföra sina LCOE och NPV.

Det här dokumentet beskriver resultatet av masterprojektet, det vill säga själva verktyget, vad det innehåller och hur det fungerar. Den metod som antagits för att designa komponenterna presenteras grundligt samt hur kostnaderna beräknades. Dokumentet förklarar de årliga prestationsberäkningarna och den enkla operationsstrategin som implementerats. Slutligen genomfördes en teknisk och kostnadsvalidering, men det skulle kräva ytterligare insats för att göra arbetet fullständigt. Konstruktionen och kostnadsberäkningarna utförs på få sekunder, de årliga beräkningarna tar cirka 2-3 timmar.

Ett huvudbidrag av examensarbetet är att visa att utformning, uppskattning av kostnader och beräkning av årliga prestanda är möjlig i ett enda verktyg som arbetar på en detaljrik nivå. Att använda verktyget under ett soltornsprojekt kan betydligt underlätta den nuvarande processen på plats hos företaget. Det kan också göra det möjligt att jämföra ett viktigt antal konfigurationer för att bestämma en bra tekno-ekonomisk lösning.

Nyckelord: Soltorn, EBSILON®Professional, Modellering, Smält salt, Designmetodik, Kostnadsberäkning, Årlig prestanda.

(4)

(5)

Master of Science Thesis EGI_2018_0101-MSC EKV1222

Development of an integrated tool to design, estimate cost and calculate annual performances of a solar power tower

Emil Blampain

Approved Examiner

Björn Laumert

Supervisor

Björn Laumert

Commissioner Contact person

Abstract

This Master Thesis consisted in realizing a tool able to design, estimate the cost and calculate annual yields of a molten salt solar power tower. Such tool was made for a company providing CSP equipment and plant solutions for engineering, engineering and procurement or also EPC of a solar power tower. The Company wishes to propose competitive plant configurations presenting a good trade-off between cost and revenues. The Company can oversee the EPC of a whole power plant or/and supply some components of the molten salt cycle and of the water/steam cycle. The tool models a large scale solar power tower with a thermal energy storage system on EBSILON®Professional 12.04, a thermodynamic software.

When launching a simulation, the tool sizes the components of the molten salt cycle (design phase) according to user’s inputs, the other components have their characteristics based on a reference project. Depending on the size of the components, the total cost is determined and the revenues over a year of operation are calculated (annual performance). When performing several simulations with different configurations, the Company can judge about the economic viability of plant configurations by comparing their LCOEs and NPVs.

The present document describes the result of the Master Thesis, that is to say the tool itself, what it contains and how it works. The methodology adopted to design the components is presented in depth, the way costs were calculated is exposed. The document explains the annual performance calculations and the simple operation strategy implemented. Finally, a technical and cost validation was carried out but it would require some further work to be complete. The design and cost calculations are performed in few seconds, the annual calculations take around 2-3h.

One main contribution of the Master Thesis is to show that designing, estimating costs and calculating annual performances is feasible in a single tool operating at a high level of detail. Using the tool during a solar power tower project could considerably facilitate the current process in place at the Company. It can also allow to compare an important number of configurations to determine a good techno-economic solution.

Key-words: Solar power tower, EBSILON®Professional, Modelling, Molten salt, Design methodology, Cost estimation, Annual performance.

(6)

(7)

ACKNOWLEDGEMENTS

I would first like to thank Nils Ahlbrink, my supervisor and head of the New Product Introduction team. He made this Master Thesis possible, I am very grateful for his regular help along the course of the project, for his many useful advice and suggestions. He also modelled the solar field and the receiver of the model and helped to debug it.

I would like to thank Julien Vauvy, Calculation Engineer in the NPI team, for being always available and ready to answer my numerous questions. All his help allowed me to move forward and he modelled the steam generator part of the system.

I express my gratitude to Arturo Aguinagua, Lead Engineer in the NPI team, for his regular follow up of my work, for modelling the water steam cycle and for his useful comments and advice.

I acknowledge Lionel Aimi, Lead Engineer at RSP, for his support. He helped me to model the annual performance strategy and allowed me to see clearer in the several cost functions I had to deal with.

I would like to thank the whole RSP division for welcoming me and integrating me nicely.

On KTH side, I am grateful for Rafael Guedez’s follow up that enabled to improve greatly this document. I also thank Björn Laumert for accepting being my supervisor.

Finally, I would like to express my gratitude to my wonderful family for their unconditional support and their help.

(8)

(9)

NOMENCLATURE

Notations

Symbol Description

𝐶 Cost (€)

𝐸 Electrical Energy (Wh)

𝑔 Gravitational constant (m/s²)

𝐶𝑝_𝑀𝑆 Specific heat capacity of molten salt (J/kg/K)

ℎ Enthalpy (J/kg)

𝐻 Head (m)

∆𝐻 Height difference (m)

𝐼𝐷 Inner diameter (m)

𝑚̇ Mass flow (kg/s)

𝑀 Mass (kg)

𝑂𝐷 Outer diameter (m)

𝑃 Pressure (Pa)

𝑃_𝑒𝑙 Electrical Power (W)

∆𝑃 Pressure drop (Pa)

𝑄̇ Thermal power (W)

𝑇 Temperature (°C)

∆𝑇 Temperature difference (°C)

𝑇ℎ Thickness (m)

𝑣 Velocity (m/s)

𝑉 Volume (m³)

𝑉̇ Volume flow (m³/h)

𝛼 Annualization factor (-)

𝛽 Discount factor (-)

𝜆 Thermal conductivity (W/m/K)

𝜌_𝑀𝑆(𝑇) Molten salt density in function of temperature (kg/m³) 𝜑 Heat transfer coefficient (W/m²/K)

(10)

Abbreviations

API American Petroleum Institute

ASME American Society of Mechanical Engineers CAPEX Capital Expenditure

CRT Central Receiver Tower

CSP Concentrated Solar Power

DNI Direct Normal Irradiance

DSG Direct Steam Generation

EHT Electric Heat Tracing

EPC Engineering, Procurement and Construction

HTF Heat Transfer Fluid

LCOE Levelized Cost Of Electricity

MSC Molten Salt Cycle

MS Molten Salt

MSCR Molten Salt Central Receiver

NPI New Product Introduction

O&M Operation and Maintenance OPEX Operational Expenditure

PB Power Block

PPA Power Purchase Agreement

SF Solar Field

SGS Steam Generation System

SM Solar Multiple

ST Steam Turbine

TES Thermal Energy Storage

TESS Thermal Energy Storage System

TMY Typical Meteorological Year

TOD Time Of Delivery

VFD Variable Frequency Drive

WSC Water/Steam Cycle

(11)

LIST OF FIGURES

Figure 1. Structure of project finance [3] ... 15

Figure 2. Annual solar energy received compared graphically to the annual energy consumption and the total resource of fossil fuels [1] ... 18

Figure 3. CSP potential worldwide [7] ... 19

Figure 4. Gemasolar power plant in Spain (19.9 MWe) [12] ... 20

Figure 5. Functioning principle of a MS solar power tower with TESS [15] ... 21

Figure 6. Sketch representing cosine, shadowing and blocking losses [1] ... 22

Figure 7. Sketches and picture of the MSCR in Solar Two [13] ... 23

Figure 8. Molten salt tank under construction [20] (left) and production of a CSP plant during peak hours [21] (right) ... 24

Figure 9. Sketch of the Steam Generation System [22] ... 25

Figure 10. Sketch of a molten salt pump for parabolic trough plants [2] ... 26

Figure 11. Example of a MS globe valve [24] ... 26

Figure 12. Pipe equipped with EHT line [14] ... 27

Figure 13. Phases of a project [3] ... 28

Figure 14. Solar power tower with its sub-systems [3] ... 29

Figure 15. Storage and turbine sized for base and peak load plants [1] ... 30

Figure 16. Influence of the storage size on the LCOE obtained with OPTYSIM, other plant variables are kept constant [30] ... 32

Figure 17. Picture of EBSILON's main window ... 34

Figure 18. Inside of a pump macro composed of fixed and variable speed pumps and their motors ... 35

Figure 19. Macro-objects on main window (left) and some inputs of the Data Center (right) ... 37

Figure 20. Picture of the thermodynamic cycle on EBSILON's main window ... 38

Figure 21. Design sequence of the model ... 40

Figure 22. Process to simulate 1 hour of the year ... 41

Figure 23. Definition of Hmax and Hmin ... 45

Figure 24. HSystem is inside the window [NStage x HMin Stage , NStage x HMax Stage] ... 46

Figure 25. Add of a stage and decrease of the speed ... 46

Figure 26. The system's head is outside the window so the rotational speed is decreased ... 47

Figure 27. Calculation of the required number of pumps ... 48

Figure 28. Cold pump station on in the thermodynamic model ... 48

Figure 29. Pump network, red crosses represents fittings with additional pressure loss ... 50

Figure 30. Three different riser configurations (1: left, 2: middle, 3: right), a change in colour indicates a change of pipe dimensions ... 51

Figure 31. EHT component simulated for a threshold of 275°C ... 52

Figure 32. Tank design sequence ... 53

Figure 33. Decision Matrix with questions and buttons to answer ... 54

Figure 34. Not-to-scale scheme of a MS pump disposed above a tank ... 55

Figure 35. Not-to-scale scheme of the bottom and lateral shell ... 56

Figure 36. Strategy to define the MSCR operating mode and target ... 60

Figure 37. Strategy to define the PB operating mode and target ... 61

Figure 38. Relative increase of stages (left) and pumps (right) given system's characteristics and based on a reference ... 64

Figure 39.Results of the function PipeDesign called in Scrip 1 ... 65

Figure 40. Results from the internal tool ... 65

Figure 41. Pressure loss evolution calculated by the tool ... 66

Figure 42. Tank levels in the morning (a), at noon (b), in the evening (c) and during the night (d) ... 67

(12)

Figure 43. Evolution of parameters during a summer and winter day ... 68

Figure 44. Evolution of the electricity production normalized and the PB operating mode ... 69

Figure 45. Evolution of the hot and cold tanks' cost in function of the storage time ... 71

Figure 46. Evolution of the plant net power and the revenues normalized ... 71

(13)

LIST OF TABLES

Table 1. Some CRT projects [16] ... 21

Table 2. Presentation of the three first project phases coming from [3] ... 28

Table 3. Molten salt properties [14] ... 43

Table 4. Example of a schedule for a given norm and material ... 50

Table 5. Definition of operating modes ... 59

Table 6. Choice of sub-profiles according to the operating modes ... 61

Table 7. Relative differences in percentage between tool results and supplier offers ... 63

Table 8. Relative difference in percent between EBSILON and the Company’s internal tool ... 66

Table 9. Relative cost difference between the EBSION tool and the Reference Plant for some components in percent ... 70

Table 10. Values from riser's configuration 2 and 3 divided with values of configuration 1 ... 72

(14)

1 INTRODUCTION

Climate change, pollution, population growth push institutions, policy makers and stakeholders to foster the production of electricity coming from sustainable energies. The sun, by providing yearly 6 200 times humankind’s annual consumption of energy [1], is an incredible resource of renewable energy that will be increasingly used in the future.

There are two main paths to produce electricity from solar energy. The first option exploits the photoelectric phenomenon to convert a share of the incident solar radiation on a semiconductor into electricity. This phenomenon is used on photovoltaic panels. In the second option, sunbeams reaching a collector (optical mirrors) are concentrated onto a receiver. In the receiver, solar energy is converted into thermal energy and transferred to a HTF. The fluid then discharges this energy in a power block where electricity is produced. This process is realized in a Concentrating Solar Power (CSP) plant.

Solar Power Tower is a promising CSP technology that is becoming mature. The good solar-to- electricity efficiency (14-18%, [2]) and the possibility to integrate a TESS are some key advantages. However, despite this positive trend, the technology is too expensive nowadays, with a LCOE between US$ 0.17-0.29/kWh [2] (this value can vary a lot between projects and locations).

These high figures foster Solar Power Towers EPC companies to build power plants with competitive characteristics. With this intention, they have to design, estimate the cost and calculate the annual performances of different plant configurations to compare their financial values and find out the best compromise between cost and revenues. The comparison of financial values will inform about the profitability of each configuration and the Company will select the most interesting solution.

The Master Thesis was carried out in Belfort (France) at Renewable Steam Plants, (named the Company in this document) part of GE Renewable Energy. The Company provides CSP equipment and plant solutions for engineering, engineering and procurement or also EPC of solar power towers. The Company uses internal tools to design, estimate the cost and calculate the annual performances of a solar power tower but the process is not flexible enough and doesn’t allow to compare many different configurations in detail with reasonable time frame. The development of a single tool that would perform the same actions than the current tools without their drawbacks is the subject of this Master Thesis.

(17)

1.1 Background

This part aims to present the typical design process of a Solar Power Tower. First, it is important to visualize where an EPC company (or EPC contractor) stands in relation to other stakeholders of a CRT project.

Figure 1. Structure of project finance [3]

All the actors gravitate around the Project Company or Special Vehicle Purpose (SPV). This entity is the main driver of the project. It receives the investments from the sponsors and credits from the banks and use it to pay the other actors. The Project Company buy environmental and political permits [4]. It invests in a suitable land to build the power plant. It establishes various agreements, notably the PPA with an Energy Company which will define the commercial terms conditioning the sale of electricity. It also selects the EPC contractor and an EPC agreement is established. The EPC contractor oversees the engineering, procurement and construction of the solar power plant according to the requirements of the SPV. The relations between the entities being exposed, the focus is now put on the typical design methodology for a solar power tower.

Depending on the Project Company, the EPC contractor will have to cope with more or less stringent conditions on the design. On the one hand, if all the characteristics of the plant are set by the SPV and not changeable, the EPC contractor will have a low degree of latitude and will respect the specifications. On the other hand, the SPV could expect advice from the EPC contractor on a good techno-economic solution. This option gives freedom to the EPC company on the plant’s configuration which has the opportunity to bring more value to the project.

In the second option, in case the Company acts as an EPC contractor and must propose a configuration, the following design methodology is currently applied:

1. Establishment of the basic configuration and estimation of the cost

The SPV, being not stringent, it provides the minimum inputs. Knowing the location enables to get weather data (DNI, ambient temperature, wind velocity, humidity). These elements allow to calculate the approximative annual performances of different configurations of CRTs with an existing internal tool. These pre-selected configurations vary from one another in terms of main components’ sizes (SF size, MSCR size, …). With high level cost functions, the financial values (CAPEX, OPEX, LCOE, NPV) of each configuration are determined. The Company selects the most economically interesting configuration. To summarize, the comparisons are very global and made only on few components.

(18)

2. Internal design of equipment

The first step leads to the establishment of heat balances for each component of the cycle. The components are then designed. This process is iterative since all the components are linked and interacts together. In the current method, the components are sized with independent tools (i.e. one tool for pumps, another one to define pipe dimensions, a third one to calculate pipe pressure loss…) and the iterations are done “manually”.

3. Send of specifications to suppliers

The components being designed, specifications are sent to the suppliers for a request for quotations. The suppliers will reply by proposing components that will or not meet the requirements. It will also propose a price and a time schedule for the supply. Some technical requirements from the Company are maybe not feasible, the suppliers would correct the specifications which might impact the design. The Company would have to go back to the second step to change the design.

4. Calculation of the cost and the annual performances

At this step, the configuration is fully known. Suppliers were chosen, their answers informed on the costs and the technical parameters of the equipment. Even though there still might be some uncertainties on costs, the estimation is precise and the final annual performance calculations are performed.

The current design methodology of the Company presents some drawbacks listed hereafter:

- The choice of configuration is limited to some components and done at a high (global) level. Indeed, the Company makes vary only few parameters.

- It is currently not possible to work at a more detailed level and to compare a huge number of configurations.

- A configuration can change during the design. Changing one key component will impact the design of the other components. The design of each component is done with individual tools, therefore resizing one component implies to resize all the components, the current process is thus time-consuming.

1.2 Objectives and Methodology

The Company expresses the need to centralize the methodology in one single tool that is flexible, where the design can be easily updated and modified and that can compare a larger number of configurations at a high level of detail (low level of modelling). The comparison should be based on techno-economic values like the LCOE or the NPV.

The main objective of this Master Thesis is to develop such tool, able to design, estimate the cost and calculate the annual performances of a large scale solar power tower (i.e. output power superior at 100 MWe) using molten salt and a direct thermal storage system. The design process should be improved with focus on MSC details and more automation.

To reach this final objective, several tasks were performed:

(19)

- Create a thermodynamic cycle equivalent to a solar power tower on EBSILON®Professional. This model would be improved during the Master Thesis.

- Implement the design rules in the model by using interconnected customizable components.

- Integrate low level cost functions to the model.

- Define and implement an operation strategy of the power plant to calculate annual performances with an acceptable computation time (order of magnitude of one hour).

- Write a technical document describing the model: methodology, equations, assumptions…

The objectives were completed with the following methodology:

- First, a literature review of existing design methodologies was performed. This was completed with the review of internal documents.

- Second, a basic thermodynamic cycle was modelled. Customized components including the Company’s design rules and costs functions were added to the cycle one by one.

- Third, as a start, a simple operation strategy was defined and implemented in the software.

- Finally, the tool was tested to validate it both technically and economically.

1.3 Thesis structure

The document will first present what motivates the Master Thesis, the objectives and the limitations. Then will follow a literature overview including an assessment of the solar resource, a state-of-the art of the CRT technology, the review of modelling and design methodologies.

The third chapter, is the core part of the Master Thesis and will detail the tool construction. It will start with a presentation of the EBSILON®Professional software and its potential. Then will be presented the tool and its functioning principle. Knowing the software and the tool, the lector will be able to understand the modelling methodology that is to say how components were designed, the assumptions used, etc. The document will expose how costs were determined and a cost function will be given as an example. The third chapter will end with a description of the annual performance calculations and the operation strategy adopted.

The chapter 4 is dedicated to the validation of the tool. The results of the tool will be compared to values of a Reference Plant designed by the Company. The validation will cover technical and cost aspects. This will lead to discussions.

Finally, the lector will discover the conclusions and recommendations to improve the model.

1.4 Limitations

Before moving forward with the Master Thesis, some limitations must be borne in mind.

The tool concerns large scale (>100 MWe) CRT using MS as a HTF and with direct storage, there is no auxiliary burner. The power block is a Rankine cycle with reheater. The design methodology is not adapted to smaller power plants.

In the present case, the Company is also providing the MSC and the WSC. It was chosen to design only the MSC (pumps, pipes, TES, SGS). The sub-systems MSCR, SF and PB are modelled according to suppliers offers from a Reference Project but their design is not changed to limit efforts given the time available. The different configurations that can be tested are thus MSC configurations.

(20)

2 LITERATURE OVERVIEW

This first chapter intends to give the lector some knowledge about the solar resource available on Earth, the functioning principle of a CRT followed by an overview of modelling and design methodologies currently existing.

2.1 Solar Resource

As stated in the introduction, the sun is an incredible source of energy. It provides to the Earth’s surface 8.85 10¹¹ GWh every year which represents 6 200 times humankind’s annual consumption of commercial primary energy in 2008 [1]. The figure below represents graphically this tremendous amount of energy alongside with fossil fuel reserves.

Figure 2. Annual solar energy received compared graphically to the annual energy consumption and the total resource of fossil fuels [1]

Besides being thousand times bigger than the annual consumption it also outweighs the cumulated energy represented by fossil fuels reserves. Wind, hydro and photosynthesis derives from the solar energy and represent a small amount compared to the sunbeam’s energy. The gravitational and geothermal energy that do not derive from the sun are not represented.

What can be concluded is that the solar energy is an abundant clean energy that should be increasingly used to ensure sustainable development of the societies.

At the Earth’s distance from the sun, the power of sun rays is about 1368 W per square meters.

During the day, 51% of this power reaches the ground on an average [5]. The rest of the incoming solar rays are mostly reflected and absorbed by the atmosphere and the cloud.

The global irradiation (power per unit area) reaching the Earth’s surface can be split in two parts:

beam and diffuse radiations. The beam radiation is the direct radiation coming from the sun, to understand it simply, one may remember that it is at the origin of shadow appearing behind objects.

Diffuse radiation is the consequence of previous reflections [1]. In the case of CSP, only beam (direct) radiation can be concentrated onto a receiver contrary to photovoltaic panels that can produce electricity from both components of the global irradiation.

The solar resource is obviously not constant throughout a year and during the day. The global irradiance is on average higher during summer than during winter. In a day, clouds can

(21)

significantly decrease the global irradiation and it is possible to have no beam radiations during an entire day which would lead to no power production for a CSP plant.

For being competitive, literature agrees on a minimum average resource of 2000 kWh/m²/year at the place where to build a CSP plant [4] [6]. This criterion is fulfilled at locations in colour on Figure 3.

Figure 3. CSP potential worldwide [7]

Suitable locations are southwestern North America, South America, North Africa, southern Africa, Middle East, China and Australia. It usually corresponds to latitudes from 15° to 40° North or South. Arid or semi-arid deserts with a good DNI resource across the year are usually places of choice. Latitudes close to the equator, though being hot, usually offer a poor DNI during summer because of the humidity [6].

Despite the previously mentioned restrictions (no diffuse radiation, limited areas), CSP has an interesting future. IEA foresees that CSP will contribute up to 11% of the total electricity consumed by 2050 [6]. This is reinforced by [2] that determined that 4801 MWe were functioning, 2837 MWe were under construction and 8472 MWe were under development in January 2016.

Actually, locations are not lacking for the production of electricity from CSP plants but their remoteness from populated areas is a drawback. Indeed, energy is lost when transporting electricity over great distances, such phenomenon can be mitigated with the use of high voltage direct current lines [6].

2.2 Central Receiver Towers

Among the four main CSP technologies (parabolic trough, CRT, linear Fresnel reflectors and parabolic dish), Central Receiver Tower is one of the most promising with a significant room for improvement [1] [8] [9].

The first demonstrative/experimental projects appeared in the 1980s. One can name for example Solar One, Solar Two (1981, 1996, USA), Themis (1984, France) and Jülich (2009, Germany) [9].

(22)

Since then, several commercial power plants were built bringing up the installed capacity to around 497 MW in 2016 [10].

Figure 4. Gemasolar power plant in Spain (19.9 MWe) [12]

The heat transfer fluid, i.e. the fluid that will carry thermal energy form the receiver to the PB can be of different natures. Air, water/steam and molten salt were used so far. Air has the advantage of being economic and not limited by a maximum temperature. High efficiencies can be expected if couple with a Brayton-cycle [1]. However, its energy density is low and the poor thermal conductivity do not favour heat transfers in heat exchangers. Water/steam is also economic but its temperature range is limited by existing ST [11] and direct storage is only available for a short period [2]. Indirect storage could be used for these two previous HTFs but it implies heat transfer losses. Molten salt, already used in the heat-treating and industry process plants [13] has a maximum temperature of 600°C but a higher energy density and thermal conductivity [14]. It can be easily stored directly in thermal energy storage tanks which avoids heat transfer losses in heat exchangers. The document will focus on solar power towers using molten salt as HTF as it is directly linked with the Master Thesis. It will also be the most used HTF for future CRT projects.

The properties of MS are listed in detail in Section 3.3.1, it is commonly composed of 60% sodium nitrate (NaNO3) and 40% of potassium nitrate (KNO3) by weight [2].

The basic functioning principle of a MS solar power tower equipped with a thermal energy storage system will be explained with the help of Figure 5.

(23)

Figure 5. Functioning principle of a MS solar power tower with TESS [15]

When the sun is shining, heliostats (mirrors) composing a heliostat field concentrates the sun rays on the receiver located at the top of the tower. The receiver converts the solar energy into thermal energy. MS at a temperature of ~290°C flows from the cold storage tank (in blue) to the top of the tower. In the receiver, thermal energy is absorbed by the MS, rising its temperature to ~565°C.

“Hot” salt flows by gravity to the hot storage thermal tank (in red). When electricity needs to be produced, hot salt flows from the hot tank through the steam generator to the cold tank. The steam produced in the steam generator is part of a conventional Rankine cycle. The steam at the exit of the ST is cooled down using a wet or dry cooling system.

MS at a temperature of ~290°C will be named cold salt in the rest of the document. MS at a temperature of ~565°C will be referred as hot salt.

The concentrating ratios at stake are important (150 to 1500 [9]). This enables a high MS temperature at the outlet of the receiver. As the temperature improves the Carnot efficiency, the overall solar-to-electricity efficiency of the power plant is good. Some solar power tower projects using MS and direct storage are presented below.

Table 1. Some CRT projects [16]

Project Power Storage Location Commissioning year Gemasolar 19,9 MWe 15h, 2 tanks Spain 2011

Crescent Dunes 110 MWe 10h, 2 tanks USA 2015

Yumen 50 MWe 9h, 2 tanks China 2018 (under construction)

The basic principle of a solar power tower being exposed, main parts (SF, MSCR, TESS, MS cycle) will be presented. It was chosen to not present in detail the PB (classical Rankine water/steam cycle) that is not proper to a solar power tower and can be found in any thermal plant with a ST. The possibility to have an auxiliary burner to enhance production, dispatchability and capacity factor is not investigated in this Master Thesis.

(24)

Solar field

The solar field is made of individual mirrors with a size varying from 1 m² to 150 m² [4]. They rotate around two axes in order to track the sun’s course and reflect the sun rays onto the receiver that is fixed. The tracking can be done differently either with an open or closed loop system. In the open loop system, heliostats are moving according to astronomic formulas (equations giving the sun’s position depending on the day, the time, the latitude and the longitude). In the closed-loop system, the heliostats track the sun thanks to sensors. The first method is more economic on a large scale while the second one offers more accuracy but can be disturbed during cloudy days [8]. It is not necessary for the land to be completely flat unlike for parabolic troughs since the mirrors are mounted on individual infrastructures. The supporting infrastructure should be designed to resist to wind loads.

The solar energy reflected by the solar field is impacted by many factors. The reflectiveness of the mirrors is not perfect and can be lowered with dust. For this reason, mirrors must be cleaned regularly. The heliostats are not directly facing the sun rays since they should direct the solar flux on the receiver, a part of the flux is lost (cosine loss). Mirrors are also interacting together; one mirror can create shadow on the one behind (shadowing loss). When reflecting the solar rays, part of the flux can be blocked by other mirrors (blocking loss) [1].

Figure 6. Sketch representing cosine, shadowing and blocking losses [1]

In addition, spillage and attenuation losses must be considered [1]. Estimating all these losses accurately requires significant computational resources. The supplier of the SF usually provides an efficiency map that, together with the DNI, the sun angles, the tower height and the heliostats’

area enable to calculate an incident solar flux on the receiver.

The SF accounts for the biggest share in the solar power tower’s cost. The cost was estimated to be around $200/m² in 2010 [4]. For Crescent Dunes, assuming a same cost, the aperture area of 1,197,148 m² [16] would lead to a SF cost of nearly M$ 239,43. In addition, the heliostats must be cleaned regularly and account for a significant O&M cost every year.

Molten Salt Central Receiver

In such component, the incident flux from the SF is converted into heat and transferred to the HTF.

It is located at the top of the tower. There are two types of receiver: the cavity-type receiver and the external-type receiver [5]. For the cavity-type receiver, the sunbeams enter in an aperture where they reflect on cavity walls and heat up the HTF. The aperture is minimized to reduce heat losses [5]. The cavity constrains the SF to 180° (except if there are several cavities) which implies a good optical efficiency at noon but a poor one in the morning and in the afternoon. External-type

(25)

receivers are more adapted for large scale power plants [4]. It is of a cylinder shape and can receive flux from any direction. Panel of tubes in which flows the HTF are disposed all around the cylinder.

The heat losses are usually bigger for such type of receiver [5].

Figure 7. Sketches and picture of the MSCR in Solar Two [13]

As can be seen in Figure 7. in the middle, the MS flows from the MSCR inlet tank (vessel) through the tubes where it is heated. It then enters the outlet tank. The vessels are acting as buffers in case of an emergency. The higher internal pressure in the inlet vessel overcomes pressure drops in tubes and the height difference to ensure the MS flow.

Based on results from Solar Two, Vogel and Kalb estimated the cost of the receiver, the tower and the MS piping in the tower of a 200 MWe CRT to be around M$ 59 in 2002 with a solar multiple (SM) of 2.7 [17]. The SM is the ratio between the collector’s peak power and the nominal thermal power of the cycle [18]. A large SM means that the SF is overdesigned and that salt must be stored.

Thermal Energy Storage System

The TESS is a very important element of a CSP plant that makes it more economically viable and increases its capacity factor and dispatchability. It will equip more than 70% of the CSP plants currently in construction [19]. They can be classified in different ways: storage duration, direct/indirect, active/passive…

Storage duration: The thermal energy can remain in the TESS for a season (store energy for the winter), a day (store energy to produce electricity during peak hours), few hours (compensate a low DNI due to clouds) [9]. Currently, TESS are mostly designed to allow production during peak hours when prices are high. Large tanks can allow to produce during 24h but then the CSP plant enters in competition with classical thermal plants (nuclear, coal) that produce cheaper electricity.

Direct/indirect storage: In direct storage, the HTF is also the storage medium whereas for indirect storage, the heat is stored in another material. In the second option, the storage material is often cheaper or has higher maximum temperature or a higher energy density. Concrete, industrial

(26)

wastes, MS (if the HTF is oil for example) can be used [19]. However, energy is lost during heat exchanges.

Active/passive storage: If the storage is indirect, an active type means that the storage material is also a fluid contrary to a passive type where a solid material is used.

The heat is usually stored as sensible heat and it is supposed to remain the case for future applications [19]. In this Master Thesis, two direct MS storage tanks were considered (see Figure 5.). They represent huge infrastructures and are well insulated to limit thermal losses with the surrounding air. The TESS represents around 14% of the investment cost according to [2].

On the left of Figure 8., one can notice the size of one tank. The graph on the right justifies the use of TES. The solar field energy does not coincide with the high energy prices when the demand is important. With the TESS, the plant can produce during peak hours i.e. in the morning and in the evening.

MS Cycle

In this part will be gathered the rest of the CRT components of the MS cycle. Components part of the MS cycle are components in which flows MS. It is the case for the MSCR and the TESS previously mentioned. But this loop also includes pumps, pipes, valves and the SGS.

Steam Generator System

In the SGS occurs the heat transfer between the MS and the water/steam of a Rankine cycle. The two fluids enter the heat exchangers with opposite flows. It is usually composed of an economizer, evaporator, superheater and reheater as represented in Figure 9.

Figure 8. Molten salt tank under construction [20] (left) and production of a CSP plant during peak hours [21]

(right)

(27)

Figure 9. Sketch of the Steam Generation System [22]

The salt temperature drops progressively from ~565°C to ~290°C across the various heat exchangers. The outlet steam pressure and temperature are set by the ST characteristics. For the Solar Two projects, shell and tube heat exchangers were used with MS on the shell side and water/steam on the tube side [14].

Pumps

The molten salt pumps move the HTF through the different components of the cycle. They are thus very important. Pumps are at least required to bring the cold salt to the top of the tower (blue path from cold tank to receiver in Figure 5.) and should overcome the height difference, pressure discharge in pipes and a pressure difference between the inlet vessel tank and the cold tank (usually at an atmospheric pressure). Pumps are also needed to move hot salt through the SGS and should overcome a height difference if any and pressure losses caused by pipes and heat exchangers.

Variable multistage vertical turbine pumps are usually used [14]. They are located above the cold and hot tanks and pump the salt from them. Their shaft must enable the suction of salt at the bottom of the tanks. That is why the shafts must be long and their length is around 15m [23]. The rotor and stator are located at the bottom of the tank and a submerging level must be ensured at any moment. The salt is leaving the pump horizontally.

(28)

Figure 10. Sketch of a molten salt pump for parabolic trough plants [2]

Valves

Valves allow to monitor the MS cycle. They are used for throttling and control of the flow. For these applications, the valves usually used are globe valves. Gate valves are utilized for isolating, venting and draining applications [14]. Globe valves are manufactured for high flows and high HTF temperatures. Valves are insulated to limit heat losses. Typically, they have a coating of stellite alloy to avoid corrosion because a corrosion layer could prevent a proper closing or opening of the valve [14].

Figure 11. Example of a MS globe valve [24]

In Figure 11, the plug (purple part) moves vertically and stops the flow when it is in contact with the seat.

Pipes

Pipes link the components of the cycle together. They have different inner diameters depending on the mass flows. Their materials and thicknesses depend on the pressures and temperatures of

(29)

the MS. They are made in steel and insulated. They can represent a significant share of the cost of the MS cycle.

Electric Heat Tracing

The EHT lines located along the pipes but under the insulation can provide heat to prevent the MS from freezing. It can be used to pre-heat an empty pipe that has cooled down, it also protects the equipment from extreme thermal stress. Valves, heat exchangers and pump shafts are also equipped with such device [14]. The heat comes from the Joule heating phenomenon. Usually, several additional lines are disposed along the component to ensure redundancy.

Figure 12. Pipe equipped with EHT line [14]

On the figure above, the EHT line (in black) is effectively disposed in contact with the pipe and under the insulation.

This sub-chapter gave the lector some first knowledge about the main parts of a MS solar power tower which will help to understand further the methodology in Section 3.

2.3 Modelling a CSP plant

The attempts to model solar power towers are numerous in the literature, some examples can be found in: [9] [25] [26] [27]. They differ from one another in terms of modelling level (the CSP plants are modelled more or less in detail), accuracy, software or coding language used, etc. The models were developed to answer a certain research question but were not suitable for this Master Thesis in which must be developed a model proper to the Company, simulating at a high level of detail. Very few documents providing guidelines on how to model a CSP plant were found. The work realized by Hirsch [3] is the only document that exposes a standard methodology for modelling such a power plant for annual performance calculations.

Modelling accurately a thermal solar power plant both technically and economically is of primary importance to provide information for the stakeholders on the viability of a project. Before

(30)

financing a project, sponsors and banks should have a precise idea of the cost and the revenues generated by the future power plant. Assessing revenues implies to calculate the energy yield from a given model on an annual basis.

But the modelling task is challenging and is impacted by three main uncertainties as described in [3]:

• Modelling approach: the mathematical models representing the components might be approximate and simplified.

• Technical parameters: the inputs of the mathematical models might not be the same than the future real components.

• Boundary conditions of the simulation, e.g. the DNI, can also be impacted by many uncertainties.

The objective is to keep the uncertainties as low as possible. The earlier the phase of the project, the bigger the uncertainty. This is because in the course of a project, there is progress and decisions are made which tend to set the configuration of a plant. SolarPACES distinguishes 4 phases of a project represented below.

Figure 13. Phases of a project [3]

The three first phases are described on the table below. The 4^th phase comes after the construction of the plant and is not treated in this Master Thesis.

Table 2. Presentation of the three first project phases coming from [3]

(31)

Hirsch also defines 2 modelling approaches. The granulated approach models the system thanks to physical equations that describes the phenomena at stake. It leads to a large number of equations and is quite detailed. The integral approach relies on empirical relations based on experiments.

This method goes less in depth and requires values of similar existing plants or components [3].

Technical modelling Sub-Systems

SolarPACES suggests splitting the model in several sub-systems. The advantages are that the complexity of a sub-system is lower than the whole system and every sub-system can be modelled independently by experts. The sub-systems are linked together at a system level. The report identifies 5 sub-systems for a CSP plant [3]:

- Solar Field (SF): heliostat field and receiver.

- Power Block (PB): MS/water heat exchangers, water/steam cycle and generator.

- Thermal Storage (TES): storage system and heat exchangers if any (not the case for direct storage).

- Auxiliary Heater (AH): fossil burner.

- Electrical system (EL): balance unit for produced and consumed electricity including transmission and transformation losses.

For each sub-system, modelling guidelines are given. For example, for the TES, useful variables are exposed, equations modelling the energy content, charging and discharging process are given.

The report also lists thermal losses that should be considered [3].

Figure 14. Solar power tower with its sub-systems [3]

Sub-systems, represented on Figure 14., need to be connected at a system level since they influence each other’s. The lines linking the components are pipes, they are modelled in the sub-system they belong to. Pipes at a system level are part of the Solar Field sub-system. The pumps are modelled at a system level since they link at least two different sub-systems, the pressure ∆𝑝_{𝑝𝑢𝑚𝑝} they have to deliver is given by Equation (1).

(32)

∆𝑝_{𝑝𝑢𝑚𝑝} = ∑ ∆𝑝𝑗 𝑠𝑢𝑏−𝑠𝑦𝑠𝑡𝑒𝑚 𝑗 (1)

The maximum possible pressure difference is selected, named ∆𝑝𝑠𝑢𝑏−𝑠𝑦𝑠𝑡𝑒𝑚 𝑗 for each sub-system.

The heat and mass conservations must be ensured at a system level. The operation strategy is modelled at a system level as well.

Annual-performance

Hirsch advices to base the energy yield calculations on a period of one year with a 10 min long time-step for accurate results, a time-step of 1h is acceptable at pre-feasibility phase. Data coming from a Typical Meteorological Year are sufficient [3].

The operation strategy defines rules and actions on how to operate the plant depending on boundary conditions (DNI, electricity prices…) and the state of the plant’s parameters (level of salt in tanks…). It influences the electricity production and the revenues. It should be defined at early phases of the project since it is related to the component design (e.g. turbine and storage sizes) and can be adapted to the location where the market sets the electricity prices.

Figure 15. Storage and turbine sized for base and peak load plants [1]

A base load operation strategy will lead to a plant with a smaller turbine compared to a plant with a peak load operation strategy for approximately the same energy output (Figure 15.).

The report defines three levels of operation strategies. The first level regroups simple and straightforward operation strategies often used as reference such as the “solar driven” strategy.

Typically, with this strategy, the PB produces electricity as soon as the DNI is sufficient. If the PB is running at full load and there is an excess of solar power, the storage tank is charged. It is discharged to keep the PB running at full load once the DNI decreases until emptiness of the tank [28]. The second level encompasses more complex strategies that require more coding effort.

These strategies are project specific, based on more variables (exhaustive state of the plant, electricity prices) but do not use forecasting. The third level includes “complex” strategies that rely on weather and price forecasts notably, and take decisions in order to maximize of minimize a function (e.g. maximize revenues) [3]. For instance, Casella et al. [29] exposes a function to be minimized which would lead to increased revenues.

(33)

Garcia-Barbarena and Erdocia [28], elaborated a “flexible strategy” that relies on a limited number of parameters common to all CRT plants (7 parameters, 3 to control the TES charging and the operating mode of the power plant, 3 to control the TES discharging and one to control the burner if any). Depending on the values given to the parameters, it is possible to make any kind of operation strategy from a simple one to a third level strategy.

The operation strategies can have different purposes like maximizing revenues or diminishing the number of ST start-ups (increase the number of start-ups decreases the ST lifetime). With this intention [28] elaborated two strategies from the “flexible strategy”; One increased the incomes by 1.3%, the other one reduced the number of ST stops by 67%.

Cost modelling

Technical aspects cannot be disconnected to cost values when it comes to judge about the performances of a renewable power plant. Indeed, better technical performances usually come at a higher cost.

Figures of interest are the CAPEX, the OPEX, the LCOE and the NPV. These figures will motivate or not investments in power plants.

Capital Expenditure

The CAPEX regroups all the costs linked to the engineering, procurement and construction of the power plant. It also includes the cost of the land, of utilities connections and some other additional fixed costs [3]. It can be expressed by the following formula [18]:

𝐶𝐴𝑃𝐸𝑋 = ∑ 𝐶_{𝑒𝑞𝑢𝑖𝑝𝑒𝑚𝑒𝑛𝑡}+ 𝐶𝑖𝑛𝑠𝑡𝑎𝑙𝑙𝑎𝑡𝑖𝑜𝑛 + 𝐶𝑒𝑛𝑔𝑖𝑛𝑒𝑒𝑟𝑖𝑛𝑔+ 𝐶𝑐𝑜𝑛𝑡𝑖𝑛𝑔𝑒𝑛𝑐𝑦+ 𝐶_{𝑙𝑎𝑛𝑑}+ 𝐶_𝑡𝑎𝑥 (2)

The first three terms represent the work typically realized by an EPC company. They are not exactly equal to the cost of equipment, installation and engineering payed by the EPC company since it applies a cost-to-price factor to ensure its margin. According to [3], parts of the CAPEX (equipment, engineering, installation) can be determined at a sub-system level.

Operation Expenditure

The OPEX regroups the operation and maintenance costs for a year. Equipment must be maintained, some parts are replaced, mirrors have to be cleaned, personnel must be payed… It is summarized in Equation (3) adapted from [18].

𝑂𝑃𝐸𝑋 = ∑(𝐶_𝑒𝑞𝑝. 𝑓_𝑜&𝑀) + 𝐶_{𝑙𝑎𝑏𝑜𝑟} (3)

The cost to maintain and operate an equipment is a fraction 𝑓_𝑜&𝑀 of its cost 𝐶_𝑒𝑞𝑝. 𝐶_{𝑙𝑎𝑏𝑜𝑟} is the annual cost represented by personnel.

LCOE and NPV

The Levelized Cost of Electricity and the Net Present Value are financial values that are calculated once a full year has been simulated. They depend on the capital investment but also on the performance of the CRT (electricity produced or revenues generated). They are well-suited to take techno-economic decisions on the design configuration.

(34)

The LCOE is the price at which the power plant should sell the electricity to overcome capital, O&M and decommissioning at the end of its lifetime. Equation (4) [18] gives its expression. Its value is in €/kWh here.

𝐿𝐶𝑂𝐸 =^{𝛼.𝐶𝐴𝑃𝐸𝑋+𝐶}^𝑂&𝑀^+𝛽𝐶^𝑑𝑒𝑐

𝐸_𝑛𝑒𝑡 (4)

Where 𝛼 is the annualization factor, CAPEX the capital investment, 𝐶_𝑂&𝑀 the cost for O&M, 𝛽 the discount factor, 𝐶_𝑑𝑒𝑐 the decommissioning cost and 𝐸_𝑛𝑒𝑡 the total net electricity produced over the complete lifetime of the plant. The lower the LCOE, the better. A designed configuration 1 with a lower LCOE than a configuration 2, can produce electricity at a lower price and still break even at the end of its lifetime.

The Net Present Value is the difference between the present values of the cash outflows and inflows (Equation (5) [18]). Its value is in € here.

𝑁𝑃𝑉 = −𝐶𝐴𝑃𝐸𝑋 ± ∑ ^𝐶^𝑘

(1+𝑖)^𝑘

𝑛𝑘=1 (5)

𝐶_𝑘 is a cash inflow or outflow at the year 𝑘, 𝑖 is the real debt interest and n the lifetime of the power plant in years. The interest debt lowers future cash inflows as they worth less than present cash inflows because of risks, inflation, etc. The higher the NPV, the better. A designed configuration 1 with a higher NPV than a configuration 2 should be more profitable.

Decisions on a good techno-economic solution should be taken based on these financial values that include the CAPEX and the OPEX.

Optimization

Finding the best techno-economic solution/configuration is the fruit of an optimization. The optimization consists in finding the extremums of certain functions while varying the CRT’s parameters.

Morin [30] suggests minimizing the LCOE function. Besides the LCOE, minimizing the ratio between cost and revenue is another option according to [31].

Figure 16. Influence of the storage size on the LCOE obtained with OPTYSIM, other plant variables are kept constant [30]

(35)

Figure 16. shows the influence of the storage size on the LCOE with the presence of a minimum.

It is an example; the study was made on a parabolic trough plant but the principle is the same for any CSP plant. Increase the storage while staying under a certain threshold increases the production of the CSP plant and increases the CAPEX of the tank, the overall trend is a decrease of the LCOE. Passed the threshold, since the other variables are fixed, a gain in storage size has no impact on the production since there is already enough salt to absorb the energy from the solar field. The CAPEX increases but not the production, hence and overall increase of the LCOE.

Although it is well recognized by financial institutions, one should be cautious with the LCOE, indeed this parameter do not consider the price varying effect of the electricity throughout the year and the day. Selling electricity when the demand is high leads to more revenues. Dispatchability is a key advantage of a solar power tower plant with TES and financial stakeholders should take it into account. That is why Dowling et al. [32], advice to focus on the NPV where revenues are considered. The optimisation would then maximize the NPV. Other functions including present values of cost and present values of revenues are suitable.

2.4 Designing a CSP plant

This Master Thesis is not just about modelling a power plant it is also about designing it, that is to say define the “technical parameters” (Section 2.4) that will feed the model.

The first step for designing a power plant is to define basic features of the CRT such as the power output and operating temperatures of the receiver. The HTF should also be defined as well as the solar multiple and the operation strategy [4]. In this Master Thesis, the focus is on CRT with MS and a TESS, so some values are set by default like the HTF and the operating temperatures.

After this first step, one can go deeper in the design of a configuration. Sandia National Laboratories’ reports [13] [14] [15] provide many useful information on the design of solar power towers with direct MS storage. The information are based on the Solar Two project (10 MWe, USA). [14] defines the sizing of a CSP plant as “an iterative design process lead by the project integrator with support from the collector field technical specialists, receiver engineer/ designer, and the turbine-generator manufacturer.”. The notion of iteration is very important since the components are linked in the cycle.

[14] gives detailed criteria upon which can be fulfilled a design. The power plant is divided in sub- systems: Collector System, Receiver System, Steam Generation System, Thermal Storage System, Master Control System, Electric Heat Tracing System, Electric Power Generation System, Balance of Plant. The Master Control System enables controlling and monitoring of the plant. The Balance of Plant supports all the other systems of a power plant.

For every sub-system, [14] provides guidelines and a methodology. It lists the norms to be used (API 650 for sizing tanks, ASME B31.1 for pipes…), physical phenomena to take into account, some design values (e.g. temperature threshold to switch on EHT…). The information are numerous and the next section of the Master Thesis is greatly based on this report.

(36)

3 TOOL DEVELOPMENT

This chapter describes the methodology followed to realize the tool. It begins with a description of the software used for the development. Then the tool is presented followed by a detailed description of the methodology adopted to design the components. Information are given on the cost estimation and the operation strategy is described in detail.

3.1 EBSILON

^®

Professional

EBSILON®Professional is a software developed by STEAG Energy Services GmbH. EBSILON is the abbreviation for “Energy Balance and SImulation of the LOad response of power generating or process controlling Network structures” [33]. It is used to simulate thermodynamic processes, for engineering, design and optimization of thermal plants. The version 12.04 was used.

It includes 125 prefabricated components which specification-values are defined by the user.

Specification-values are inputs (e.g. efficiency, geometrical values) to the components that will enable the thermodynamic calculation. The inputs can be reals or more complicate mathematical objects like characteristic lines or matrices.

A cycle is then built on a graphical window by disposing the components and by linking them with lines. These lines can represent fluids such as water, steam, thermoliquids or electric lines, logic lines… The picture below shows an example of a CSP plant with parabolic trough proposed by EBSILON®Professional.

Figure 17. Picture of EBSILON's main window

(37)

In Figure 17., one can identify ready-made components such as parabolic troughs on the left, turbine stages in the upper right corner, storage tanks in the middle… The grey lines represent the heat transfer fluid, the red lines represent steam and the blue lines represent water.

For the simulation, the process’ values and the equations are stored in a computation matrix that is linearized at the start of the simulation and then solved with iterations [33]. All the equations are solved assuming a steady-state.

For a given model, one can create different “profiles” for the simulation. “Profiles” differ from one another in the sense that components’ inputs can be different. There is a hierarchy between the profiles. The “Parent profile” is the design mode in which the components’ nominal values are determined. “Sub-profiles” can then be created to run the cycle in different situations (e.g. part load). The calculation of the annual performances will for example be done in a “Sub-profile” once the plant has been designed in the “Parent profile”.

Ebscript

The software offers the possibility to execute user defined scripts called Ebscripts. The programming language is Pascal-based. Through Ebscripts it is directly possible to access the components and modify them. The language is object-oriented, components are objects.

It is possible to simulate the cycle with the function Simulate(), switch between sub-profiles…

These Ebscripts can be executed before, during and after EBSILON’s thermodynamic simulation.

Macro-object

The software proposes 125 prefabricated components (pump, pipe, valve, turbine, heat exchanger, etc.). In addition, EBSILON®Professional offers to create macro-objects that are customizable.

One can define all the inputs and outputs as well as the Ebscripts attached to the macro to be run before or after the thermodynamic simulation. A macro can contain matrices, characteristic lines...

Inside the macro, a new graphical window is open and can include a whole process composed of standard components or other macros. This object allows the user to create a complicate and fully customized component.

Figure 18. Inside of a pump macro composed of fixed and variable speed pumps and their motors

If a system includes several macros with attached Ebscripts, one can decide the order of execution of the Ebscripts, indeed the design of one component done in a Ebscript can impact the design of another component.

(38)

Time series

Time series calculations enable to perform a series of calculations over a certain range of time divided in time steps. At each time steps, the software offers the possibility to run a script. This is particularly interesting for the annual performance calculation where the simulations must be run every time step (e.g. 10 min or 1 h) of the year with changing weather conditions. A script can for example apply the operation strategy and simulate the cycle accordingly. In this mode, evolution of the tanks’ volumes is a transient process because it is time dependant.

Why choose EBSILON®Professional?

EBSILON®Professional was chosen because of different reasons. First, there is a broad range of pre-fabricated components that can be customized using macros and Ebscripts. In addition, it includes all the thermodynamic functions needed for a simulation. Second, the time-series option allows to perform annual performance calculations. Third, the software can return excel sheets that would facilitate the creation of specifications for suppliers. Finally, the software includes the possibility to run optimizations with the add-on module EbsOptimize.

Other software are specifically dedicated to design, estimate the cost and calculate annual performances of CSP plants like the System Advisory Model developed by the National Renewable Energy Laboratory (USA). It can also perform optimizations. Although the solar field and the storage system are well simulated, characteristic curves are used to represent the PB [30].

Such model for the PB is not suitable in the Master Thesis where a detailed modelling is required.

EBSILON has already been used for simulating annual performances of a CSP plant [27].

3.2 Tool Presentation

The Section 3.2 intends to explain what the tool is made of, together with the processes at stake to design and calculate the annual performances of the CRT.

3.2.1 Thermodynamic model

The thermodynamic cycle is presented on Figure 20, the CRT plant is represented at a system level. One may identify the sun, the Heliostat field, the MSCR, the hot, cold and drain tanks as well as the SGS and the PB. The functioning principle of the model is the same than a classical solar power tower described in Section 2, Figure 5.

Some more information is given on the macro-objects on the left of the Figure 19. (Data Center, Design Center and Decision Matrix) and on the solar part (sun component, SF and MSCR).

Data Center, Design Center and Decision Matrix

The “Data Center” and the “Design Center” or “Calc Center” are macro-objects. The Data Center gathers all the necessary inputs that will allow to perform the design. The user fills the inputs with the desired values. There are more than 100 inputs with relative importance, for instance, the user can define the storage time, the piping material, the type of pumps (variable speed vs. fixed speed)… The Calc Center is not accessible by the user. It includes two scripts, one to be run before

Development of an integrated tool to design, estimate cost and calculate annual performances of a solar power tower