compressed air energy storage - diva1214933/...degree project in technology, first cycle, 15 credits...

IN DEGREE PROJECT TECHNOLOGY,FIRST CYCLE, 15 CREDITS

, STOCKHOLM SWEDEN 2018

Compressed air energy storageProcess review and case study of small scale compressed air energy storage aimed at residential buildings

EVELINA STEEN

MALIN TORESTAM

KTH ROYAL INSTITUTE OF TECHNOLOGYSCHOOL OF ARCHITECTURE AND THE BUILT ENVIRONMENT

1

ACKNOWLEDGMENT We would like to express our gratitude to our supervisor Assist. Prof. Justin Ning-‐Wei Chiu, for without his advice and guidance the making of this report would not have been possible. Thank you!

2

INDEX OF FIGURES FIGURE 1. SCHEMATIC IMAGE OF CAES SYSTEM. NOTE THAT THERMAL STORAGE IS OPTIONAL AS IS THE NUMBER OF COMPRESSORS

AND TURBINES. ........................................................................................................................................... 12 FIGURE 2. DETAILED DESCRIPTION OF EQUATIONS USED FOR DECIDING CHANGES IN TEMPERATURE, PRESSURE AND VOLUME FOR ALL 22

DIFFERENT STAGES IN THE CAES PROCESS. ISENTROPIC PROCESS RELATION IS ABBREVIATED AS IPE AND IDEAL GAS LAW AS IGL, BOTH USED TO DENOTE HOW THE VALUE OF THE CONCERNED PROPERTY IS DERIVED. ..................................................... 18

FIGURE 3. LOAD PROFILES FOR ONE HOUSEHOLD DURING MONDAY-‐FRIDAY AND SATURDAY-‐SUNDAY. ...................................... 23 FIGURE 4. THE BLUE AND ORANGE STAPLES SHOW VARIATIONS IN PRESSURE DURING CHARGE PHASE, DISCHARGE PHASE AND THE GREY

LINES FOLLOWS THE PRESSURE IN THE STORAGE. DURING PERIODS WHERE ONLY THE GREY LINE IS PRESENT THE STORAGE HAS

BEEN EMPTIED TO MINIMUM PRESSURE AND THE SYSTEM IS AT REST. THIS FIGURE IS TRUE FOR MONDAY TO FRIDAY. ............ 24 FIGURE 5. THE BLUE AND ORANGE STAPLES SHOW VARIATIONS IN PRESSURE DURING CHARGE PHASE, DISCHARGE PHASE AND THE GREY

LINES FOLLOWS THE PRESSURE IN THE STORAGE. DURING PERIODS WHERE ONLY THE GREY LINE IS PRESENT THE STORAGE HAS

BEEN EMPTIED TO MINIMUM PRESSURE AND THE SYSTEM IS AT REST. THIS FIGURE IS TRUE FOR SATURDAY AND SUNDAY EXCEPT FOR THE SMALL DETAIL THAT FOR THE LAST HOUR OF SUNDAY THE PRESSURE INCREASES LIKE IN FIGURE 4 TO ACCOMMODATE THE

SEVEN-‐HOUR CHARGE PERIOD FOR MONDAY....................................................................................................... 24 FIGURE 6. THE BLUE AND ORANGE STAPLES SHOW THE MASS THAT NEEDS TO BE EXPANDED TO SATISFY THE DEMAND OF EACH

DISCHARGE PERIOD. THE GREY LINE IS THE DIFFERENCE BETWEEN MAXIMUM AND MINIMUM MASS. .................................. 25 FIGURE 7. AVERAGE ELECTRICITY SPOT PRICE FOR EACH HOUR DURING WEEKDAY AND WEEKDAY FOR THE OF 2017 (NORDPOOL,

2018A). ................................................................................................................................................... 26

INDEX OF TABLES TABLE 1. EFFICIENCIES FOR VARIOUS COMPONENTS AND PROCESSES. ................................................................................ 21 TABLE 2. SPECIFIC HEAT CAPACITIES USED DURING CALCULATIONS. ................................................................................... 21 TABLE 3. ECONOMIC RESULT FOR A WEEK WITH AND WITHOUT CAES. .............................................................................. 26 TABLE 4. CAPITAL COST AND CHARACTERISTICS OF MAIN COMPONENTS. ............................................................................ 26 TABLE 5. NET PRESENT VALUE FOR TWO SCENARIOS. ..................................................................................................... 27

3

NOMENCLATURE

Abbreviations CAES Compressed air energy storage EES Electrical energy storage IGL Ideal gas law IPE Isentropic process equations HPC High pressure compressor HPT High pressure turbine LPC Low pressure compressor LPT Low pressure turbine NPV Net present value RFB Redux flow batteries TES Thermal energy storage U-‐CAES Underground compressed air

energy storage UW-‐CAES Underwater compressed air

energy storage

Variables

𝐴 Area c# Specific heat capacity constant

volume c$ Specific heat capacity constant

pressure 𝐸 Energy 𝑚 Mass �̇� Mass flow 𝑀 Molar mass 𝑝 Pressure

�̇� Heat flux 𝑅 Gas constant 𝑡 Time 𝑇 Temperature 𝑣 Specific volume 𝑉 Volume 𝑤 Work

Greek symbols 𝜅 Heat capacity ratio 𝜂 Efficiency

Subscripts

1 Inlet/Before process 2 Outlet/After process 𝐺 Generator 𝑀 Mechanical 𝑇 Total 𝑎𝑡𝑚 Atmospheric conditions 𝑠 Specific 𝑆 Storage 𝐶 Compressor 𝐻 Heat 𝑁𝐺 Natural gas 𝑅𝑇 Round-‐trip

4

ABSTRACT The potential for electrical energy storage to both provide services to the electrical grid and help to better integrate renewable energies in the electrical system is promising. This report investigates one type of storage, compressed air energy storage (CAES), where energy is stored by compressing air during hours of low electricity demand and later expanding the air to generate electricity during high demand hours. To this day it exists two large plants, but small facilities have yet to be implemented, raising the question whether it could be viable to use CAES on a smaller scale as well. By creating a model of a CAES system based on the principles of thermodynamics and applying it to a hypothetical group of residences, its ability to balance daily fluctuations in electricity demand is explored. The result show that the system is able to cover some of the demand but there is no economic profit to be gained. The results of this report suggest that a CAES system of this size is not a viable option during current price market for electricity in Sweden but during other circumstances it could be relevant.

KEYWORDS: compressed air energy storage (CAES), electrical energy storage (EES), artificial air storage, thermodynamic analysis, economic evaluation

5

SAMMANFATTNING Dagens energisystem kräver vissa tjänster för att kunna behålla stabilitet och tillgodose energibehovet. Energilagring är ett sätt att förse systemet med dessa tjänster samtidigt som det också skapar möjlighet att bättre utnyttja förnyelsebara energiresurser, som vind och sol, som annars kan vara för oförutsägbara för att kunna utnyttjas maximalt. I denna studie undersöks komprimerad luft som energilagring (CAES). Sammanfattningsvis används billig elektricitet under timmar då elförbrukningen är låg för att komprimera luft och lagra denna för att sedan expandera luften igen och på så vis generera elektricitet vid behov eller då det finns ekonomisk vinstmöjlighet. CAES systemet kan vara uppbyggt och dimensionerat på flera olika sätt vilket undersöks samt beskrivs i närmare detalj. Möjligheten att använda CAES i liten skala för att tillgodose ett dagligen varierande energibehov undersöks och det utrönas ifall detta är ekonomiskt gynnsamt eller inte. Detta undersöks genom att skapa en modell över ett CAES-‐system som appliceras på energibehovet för en grupp bostäder. Resultatet visar att systemet kan täcka en del av energibehovet men ekonomisk vinning är inte möjligt. Utifrån dessa resultat konstateras att CAES i liten skala inte är ett ekonomiskt försvarbart alternativ för att täcka toppar i ett varierande energibehov vid det rådande energipriset i Sverige men under andra omständigheter skulle det kunna vara möjligt.

NYCKELORD: komprimerad luft som energilagring (CAES), elektrisk energilagring (EES), artificiell luftförvaring, termodynamisk analys, ekonomisk utvärdering

6

TABLE OF CONTENTS Acknowledgment ......................................................................................................................... 1 Index of figures ............................................................................................................................ 2 Index of tables .............................................................................................................................. 2 Nomenclature ............................................................................................................................... 3 Abstract .......................................................................................................................................... 4 Sammanfattning .......................................................................................................................... 5 1. Introduction .......................................................................................................................... 8 1.1. Purpose ....................................................................................................................................................................... 9 1.2. Objectives .................................................................................................................................................................. 9

2. Background ........................................................................................................................ 10 2.1. Electrical energy storage ................................................................................................................................ 10 2.1.1. EES in the energy system ......................................................................................................................... 10 2.1.2. Electrical energy storage solutions .................................................................................................... 11

2.2. The CAES process ................................................................................................................................................ 12 2.2.1. Compression ................................................................................................................................................... 13 2.2.2. Storage ............................................................................................................................................................... 13 2.2.3. Expansion......................................................................................................................................................... 15 2.2.4. Thermal storage............................................................................................................................................ 15

3. Methodology ...................................................................................................................... 16 3.1. Case study............................................................................................................................................................... 16 3.2. Load Profile ........................................................................................................................................................... 16 3.3. Calculations for system dimensioning ...................................................................................................... 16 3.4. Heat treatment .................................................................................................................................................... 19 3.5. Efficiencies ............................................................................................................................................................. 19 3.6. Constants ................................................................................................................................................................ 20 3.7. Economic Evaluation ........................................................................................................................................ 21 3.8. Electricity prices ................................................................................................................................................. 21 3.9. Investments ........................................................................................................................................................... 21

4. Results ................................................................................................................................. 22 4.1. System Properties ............................................................................................................................................... 22 4.1.1. Load profile ..................................................................................................................................................... 22 4.1.2. Compression and expansion work ...................................................................................................... 23 4.1.3. Charge and discharge................................................................................................................................. 23 4.1.4. Efficiency .......................................................................................................................................................... 25 4.1.5. Heat ..................................................................................................................................................................... 25

7

4.1.6. Loss due to constant volume storage ................................................................................................ 25 4.2. Financial results .................................................................................................................................................. 26

5. Discussion ........................................................................................................................... 27 5.1. Assumptions .......................................................................................................................................................... 27 5.2. Constant volume/varying pressure vs. Constant pressure/varying volume ........................... 28 5.3. Optimization ......................................................................................................................................................... 28 5.4. Profitability ........................................................................................................................................................... 29 5.5. Large scale or small scale ............................................................................................................................... 30 5.6. Environmental aspects ..................................................................................................................................... 30 5.7. Social aspects ....................................................................................................................................................... 30

6. Conclusions and recommendations ............................................................................ 31 7. References .......................................................................................................................... 32

8

1. INTRODUCTION As global warming and climate change continue to increase and make themselves known not only by their consequences but increased awareness, the interest for sustainable solutions grows rapidly. The concept of “sustainable development” is a multifaceted term, used by many and in just as many contexts. Many definitions of sustainable development are derived from the Brundtland report, which states that to make development sustainable humans need to “ensure that it meets the needs of the present without compromising the ability of future generations to meet their own needs” (UN, 1987) but this can be interpreted in many ways. While its vague definition is by some deemed problematic, the general consensus is still that sustainable development is of the greatest importance for the future of the human race and needs to be a top priority (Kuhlman and Farrington, 2010). One of the most important drivers of development is energy, which is necessary for growth on both a individual and global level (IPCC, 2016) and also part of the UN’s sustainability goals (UN, 2016). In many parts of the world there is an abundance of energy, as society has spent both time and resources in developing the techniques of harnessing energy from sources such as oil, coal and nuclear materials. With the growing climate changes it has been made obvious that the traditional ways of energy production will no longer be able to sustain the world in ways that do not risk radically changing the global ecosystem. Furthermore, the UN sustainability goals specify that energy should be both clean and affordable, a criterion that is not fulfilled by the use of fossil fuels (Stockholm Resilience Centre, 2018). Renewable energies in many forms are being developed, but just as fossil energies have a problem fulfilling the “clean” part of “clean and affordable energy”, renewables have a problem fulfilling the “affordable” part. Within this affordability spectrum falls the problematic fluctuating properties of many renewable energies as the main sources (such as sun, wind and waves) are not constant in their supply but vary with time. A proposed solution to battle the fluctuation, and thereby making renewables a more attractive and stable form of energy, is energy storage. Today, the practiced forms of electric energy storage are pumped hydroelectric storage, certain battery technologies and compressed air as energy storage (Drury et al., 2011). While these solutions all have the ability of supporting renewable energies in terms of balancing peak-‐demand and providing back-‐up reserves, they all come with their respective set of problems. Although these solutions are already being commercially implemented to some extent, further development is required if they are to be applied on a global scale. This report will focus on investigating the field of compressed air as energy storage, commonly known as CAES. The concept of CAES is to compress air in period of excess energy, and later on expand it, releasing the energy back into the grid during periods of energy shortage.

There already exists two functional CAES plants (Garvey and Pimm, 2016) which were constructed several decades ago, and as technology has since then developed, many aspects could be improved. In the process of compressing, storing and expanding the air the main emissions originate from the burning of fossil fuels to regulate temperatures, giving the environmental issues a more uniform solution by simply ensuring that the required energy comes from renewable sources. To promote an increased use of CAES, this report will instead focus on investigating the technical performance and economic viability. Since large scale CAES plants already exist, the report will, in addition to providing a detailed description of the technical process and how CAES is being used today, investigate the

9

possibility of implementing it on a smaller scale to give energy storage capacity to a group of smaller buildings or one large building.

1.1. PURPOSE Provide an overview of the technical aspects of the CAES process and analyze its viability as energy storage on a small scale.

1.2. OBJECTIVES • Outline the full process, from compression to expansion, of CAES from a technical and

engineering perspective. This includes key aspects such as efficiency, excess heat treatment and variations due to chosen storage type.

• Investigate whether CAES is a feasible option for small scale energy storage to balance daily energy demand fluctuations and increase energy independence.

• Evaluate CAES, from an environmental viewpoint, in relation to other energy storage options and identify its role in the development of renewable energies.

10

2. BACKGROUND

2.1. ELECTRICAL ENERGY STORAGE The electric system is a complex configuration of power generating units, transmission and energy users creating supply and demand within the system. To keep the system stable, equilibrium has to exist between supply and demand of electricity (Ibrahim et al., 2008). However, the electricity demand is constantly changing, both from day to day and season to season, all depending on the users need for heating, cooling, lighting etc. (Denholm et al., 2010). The integration of renewable energies with variable and unpredictable energy output, such as wind and solar power, into the grid ads yet another dimension of uncertainty making it even harder to maintain equilibrium (Ibrahim et al., 2008). The amount of electricity from renewable energy sources in the grid has increased which is an important step towards a more sustainable energy system and a way to lessen the overall dependence on fossil fuels and thereby also greenhouse gas emissions (Denholm et al., 2010, Salvini et al., 2017). To increase the chances of more integration, the unreliability of renewable energy sources must be tackled to make it easier to integrate them into the grid. Electrical energy storage (EES) could be a solution since it could be a way to regulate the electricity supply from renewables to meet the changing demand and thereby maintain equilibrium (Ibrahim et al., 2008). EES could also be a way to strengthen reliability of the existing power grid as well as boosting integration of renewables (Denholm et al., 2010).

2.1.1. EES IN THE ENERGY SYSTEM The instability of variable renewable energies can be observed in deregulated electricity markets where the electricity price can be highly volatile and change drastically during the day because of the changes in energy output from renewable energies (Bullough et al., 2004). The price of electricity is determined by a balance between demand for electricity and the supply (Nordpool, 2018b) and a variation of different power plants is used to meet the need for electricity where a baseload plant handles the constant demand (Denholm et al., 2010). To provide baseload power a technology where the output of power can be planned regardless of weather conditions have to be used and in Sweden it consists of hydropower and nuclear power (Byman, 2016). Baseload power plants are often used as much as possible and in some cases, e.g. nuclear power, there are restrictions preventing rapid changes in output power i.e. this type of power plants can produce a constant and reliable energy output but if the electricity demand spikes the output cannot be changed fast enough to meet that demand. Usually these plants have a large capital cost and low variable cost (fuel) which encourage constant usage. Consequently, other electricity production technologies have to be used for meeting variations in load. These are called load-‐following plants, and some can be classified as intermediate meaning they meet variations in load from day to day. Then there are ones that meet unforeseen peaks in demand, peaking units (Denholm et al., 2010). Since the electricity price is determined by supply and demand the off-‐peak electricity is cheaper.

Electric energy storage solutions are a possible addition to the grid which may have potential to improve the electricity system in a multiple of ways and they are a part of future sustainable energy systems (Lund and Salgi, 2009). The improvements include contributing to meet the variation in electricity demand since EES could be an alternative to better utilize baseload plant and reduce need for plants operating with less efficiency. EES can provide load leveling effect, meaning it uses electricity during off peak and storing it to later supply electricity during high peak hours. EES also has

11

the ability to provide backup in case of temporary loss of other electricity production. Intermittent power generation from renewable resources and other variations in demand also cause frequency variations in the grid, meaning there is a need for frequency regulation which EES can provide by charging or discharging in response to increase or decrease in frequency (Denholm et al., 2010, ABB, 2016).

2.1.2. ELECTRICAL ENERGY STORAGE SOLUTIONS There are a number of different technologies for storing electrical energy. Which one that should be used depends on the intended application, meaning the characteristics of different alternatives should be evaluated and compared to the needs of the system. Efficiency, lifetime, discharge time, weight and mobility are examples of characteristics that could be relevant to considered when choosing an EES technology. Categorizing the different storage technologies in terms of their function can be helpful for deciding which one to use and it is common to divide them into energy applications, i.e. medium to long-‐term storage, and power applications which requires rapid response time (Zhao et al., 2016a, Cho et al., 2015). The later application includes providing services such as frequency regulation and contingency reserves where the storage technology must be able to respond fast to the change in demand. Longer term storage could possibly also provide these services and in addition also provide load leveling (Denholm et al., 2010, Drury et al., 2011). Some electrical energy storage solutions to consider are pumped hydroelectric storage (PHS), compressed air energy storage (CAES) and different battery technologies which all possess potential to provide energy management during a longer time span (Denholm et al., 2010).

Energy is stored with pumped hydroelectric storage by using electricity during off-‐peak hours to pump water from a lower located reservoir to one located higher up. During peaking hours, the water is released from the upper reservoir to flow through turbines, generating electricity to satisfy the increasing demand (Ibrahim et al., 2008). PHS provides large storage capacity and long-‐term storage with a long lifetime and low unit cost (cost per kWh) with an efficiency of about 70 -‐ 80% (Drury et al., 2011, Molina, 2017). Another often mentioned advantage of PHS is the maturity of the technology, meaning it is readily available and already being used (Molina, 2017, Ibrahim et al., 2008). There are three plants in Sweden using the technology with effects ranging from 600 kW to 36 MW (Byman, 2016). However, PHS suffers from geographical constraints which limits the possibility of increased development, especially in developed countries. There is also a large capital cost and some environmental issues (Bullough et al., 2004, Molina, 2017).

Compressed air energy storage is the only other commercially available storage alternative, besides PHS, with very large capacity but CAES offers less geographical constraints than PHS (Bullough et al., 2004, Cho et al., 2015). Since the subject of this paper is CAES, the process is explained in more detail in the methodology, but in short CAES, like PHS, uses off-‐peak electricity to store energy but in the form of compressed air instead of water. Two plants are in existence at present time (March 2018), one in Huntorf, Germany and one in Macintosh, Alabama in the USA. CAES can be used for both small-‐scale and large-‐scale applications. The storage capacity will of course depend on the volume available for air storage which also creates geographical constraints since underground caverns can be used as storage. However, artificial storage is a possibility which eliminates geographical problems but also limits capacity to some extent as the material costs of storage become high (Ibrahim et al., 2008). Like PHS, CAES also has long lifetime and low unit cost. However, depending on the type of CAES technology being used there may be environmental issues in the form of greenhouse gas emissions.

12

The type of technology also affects the efficiency of the process which therefore range from 45 -‐ 70 % (Molina, 2017).

Batteries are another type of storage alternative and there are many different kinds. The most promising scalable technology for applications such as storing energy generated from renewable resources is redox flow batteries (RFB). RFB are suitable for large scale and long-‐term storage. Just like PHS and CAES they also have long life time and the efficiency is about 70 -‐ 85%. There are however problems with low energy density and complicated system requirements (Molina, 2017).

2.2. THE CAES PROCESS The basic principle of compressed air energy storage is very simple: compress air during periods of off-‐peak electricity and expand it during periods of peak electricity. The CAES process can be divided into three main stages: compression, air storage and expansion. Depending on the point of view and system configuration, thermal storage can also be considered a stage in the process but bear in mind that it is not critical to the process, but an addition for improving overall performance. The setup of the system used for this report is shown in figure 1, where thermal storage is included and compression/expansion performed in two stages.

For each of these stages there are different critical aspects for obtaining optimal performance of the system. An aspect that greatly defines configuration of the entire CAES system is whether it is adiabatic or diabatic. During the CAES process heat loss occurs during all stages, and minimizing it has for a long time been one of the main focuses of CAES research (Hartmann et al., 2012). During the CAES cycle heat is produced when compressing air and used when expanding it, so for an ideal process with perfect heat storage there would be no need to add heat as the same amount is needed for expansion as gained from compression. However, storing heat is difficult and expensive, which could be why the only existing plants are diabatic.

Figure 1. Schematic image of CAES system. Note that thermal storage is optional as is the number of compressors and turbines.

13

2.2.1. COMPRESSION The compressor is motor-‐driven (Huang et al., 2018) and, depending on the dimensioning of the system, have different sizes and power output and has the main objective to compress the air going into the air storage cavity. For small-‐scale CAES it is possible to use compressors which also function as generators during the discharge process (Ibrahim et al., 2008), but this results in lower efficiencies and is not common procedure.

During the compression stage the main losses are mechanical conversion loss and heat loss that occurs due to the pressure increase. Referring to the ideal gas law, there is a direct connection between pressure increase and temperature increase, which is why (depending on pressure ratio) the temperature after the compressor can be several hundred degrees Celsius. These kinds of temperatures are highly destructive for most storage cavity material, which is why the air needs to be cooled by passing through a heat exchanger before entering the air storage (Donadei and Schneider, 2016). If the CAES system uses thermal energy storage (TES), compressing the air in one stage (aptly named single-‐stage compression) requires a TES that can withstand high pressures and high temperature, making it quite a costly affair. Instead it can be more efficient using multi-‐stage compression, where the air is compressed and cooled several times before reaching the air storage. Cooling the air between compression or expansion stages is called interim cooling or heating and lessens the requirements for the TES material as the heat which is transferred from the air has a lower temperature. For each interim cooling stage added to the compression process the heat loss and material requirements will lower but the effect of each added stage will also decrease, meaning that there will be a cost-‐benefit equilibrium between the number of compressors and the gain in decreased losses. To avoid thermal degradation, (Huang et al., 2018) suggests that the temperature must not exceed 300°C and (Hartmann et al., 2012) found that two stage compression gives highest efficiency in relation to amount of stages used. Most research papers concern adiabatic and semi-‐adiabatic CAES which always include some form of heat storage, but it is of course possible to completely ignore the thermal aspects and release the heat into the ambient air. This would require no heat recuperator or TES, but the losses would be large, and efficiency be so low as not to be cost efficient. It would also require some other heat source, such as burning natural gas.

The compressor efficiency depends on the pressure ratio between inlet and outlet and with values between 70-‐90% (Garvey and Pimm, 2016). In published work the compressor efficiency is generally approximated as a fixed value, but in reality, the pressure ratio varies as more and more air is injected into the system which indicate that the efficiency also varies (Salvini et al., 2017). The fact that the pressure ratio varies during the compression (and by the same logic for the expander) creates some technical difficulties as both compressor and expander obtain optimal efficiency when the pressure ratio is constant, and a pressure regulator is needed, creating energy losses (Pimm and Garvey, 2016).

2.2.2. STORAGE The next step in the CAES cycle is the storage. The storage must able to store large quantities of air at high pressure, something that is not easily found and leads to one of CAES great weaknesses: it is highly location specific (Garvey and Pimm, 2016). Storage systems can be either isobaric or isochoric, isochoric being more common while isobaric gives better performance but is harder to achieve. There are three main ways of dividing storage types: underground, underwater and aboveground with each type having its own set of advantages and disadvantages.

14

UNDERGROUND Storing the air below ground is the most common storage type and both of the existing plants use this method. The advantages of underground storage are that it is cost efficient in relation to the high storage capacity, is protected from external impacts and has a low ecological footprint (Donadei and Schneider, 2016). There are two main disadvantages, the first being that underground storages have fixed volumes and compression and expansion process are thereby not isobaric as pressure inside the cave varies during charge and discharge, impairing overall efficiency. The other main disadvantage is that suitable underground storage can be hard to find and requires extensive geological investigation before constructing a plant.

There are five kinds of underground cavities that can be used for air storage: depleted oil and gas fields, aquifers, salt caverns, rock caverns and abandoned mines. The air in the only two existing plants, McIntosh and Huntorf plants, is stored in large salt caverns which have the benefit of the salts low reactivity with air and low pre-‐investigation work. Even if salt caverns are the only storage cavity used today, research focuses more on underground formations as the salt caverns are location specific and have already been investigated. While aquifers have been proved appropriate for storing natural gas they have high reactivity and require extensive pre-‐investigation due to its more intricate nature. The depleted oil and gas field have already been proven fit to store gas and fluids but residuals from the previous oil or gas can cause problems for CAES and no fields have been used so far. Rock caverns are similar to salt caverns but need to be sealed to minimize self-‐discharge and could be an option where salt cavern construction is not possible but the rock is hard enough to be adequate for storage. Finally abandoned mines could also be an option but just as for rock caverns eventual cracks need to be sealed and making the mine fit for CAES can be costly.

UNDERWATER A less explored storage form is underwater CAES whose characteristics are primarily defined by its isobaric properties. For underground CAES air is stored is large cavities but for underwater CAES the researched vessels are generally much smaller, flexible and made from durable fabric. As the air is stored under water the pressure is high and by using flexible vessels the volume adjusts automatically, improving overall performance. Underwater CAES has the advantages of having low cost of installation; applicable both for oceans, seas and lakes; air can be stored at hydrostatic pressure which occurs naturally. However, the energy density is lower than that of underground CAES, no assessment has been done for impact of the local marine ecosystem and just as for underground CAES it is location specific. Underwater CAES is suitable for locations where a large water depth can be found relatively close to the shore which simplifies the installation and maintenance. It is also possible to install far from shore plants but then it would also be necessary to have an offshore platform with connecting transmission lines and TES on the platform if the plant is adiabatic (Pimm and Garvey, 2016).

Most research paper concern large scale CAES (Salvini et al., 2017), but due to the low energy density implementing underwater CAES on large scale would be problematic because of the amount of anchoring weight needed. To fix the flexible vessels to the ocean floor, gravity-‐based anchors is the most viable option, but the amount of weight needed is immense ( Pimm and Garvey (2016) mention that for a plant of 1 GWh capacity the volume of concrete needed would be a quarter of a million cubic meters). Despite this difficulty and thanks to the isobaric properties there is still interest for underwater CAES, (Sheng et al., 2017) conducted a study on combining a wind plant and underwater

15

CAES to supply a small stand-‐alone island, concluding that it could indeed be a viable power supply option.

ARTIFICIAL As discussed both underground and underwater CAES have the problem of being location specific as they depend on naturally occurring geologies. A way to avoid this would be to construct artificial storages, which has been investigated by (Liu et al., 2014) among others. There are three different kinds of storage devices: gas storage pipelines, gas cylinders and storage tanks. Liu et al. (2014) conclude that gas storage pipelines normally have a lower cost, while air storage tanks are the only option without pressure constraints. Artificial storage has a high initial cost due to expensive storage materials and a lower continuous maintenance cost. Comparing artificial storage with underground storage shows that artificial CAES is more expensive, but as mentioned with the advantage of being non-‐location specific.

2.2.3. EXPANSION After the air has been compressed and stored it is time to extract the energy by passing it through the expander. Just as heat needed to be removed during the compression, heat needs to be added during the expansion. In order to not cause harm to the expander as a result of frozen particles in the air, the air needs to be kept above freezing temperatures. For an ideal process, the energy created during the compression could be used for expansion and no heat would need to be removed or added to the system. However, due to heat storage difficulties, the two existing CAES plants need to burn fuel to provide heat during the expansion. The fuel most commonly used is natural gas (Drury et al., 2011) which contributes to global warming, but as heat storage technology has developed, future plants will likely include some form of energy management where heat from the compression is used which will decrease environmental impact.

There are different kinds of turbines that can be used and an important requirement for CAES is that the turbine must be able to handle the large variations in mass flow. The range of inlet pressure and pressure ratio requirements may vary and therefore it can be useful to have multi-‐stage expansion and/or expansion valves to maximize efficiency. The expander needs to be connected to a generator in order to extract electrical energy, and this connection can also be structured in different ways (Zhao et al., 2016b).

2.2.4. THERMAL STORAGE In terms of overall efficiency, it is beneficial to store the produced heat from the compression stage to use it later on, either for expansion or other application. To increase overall efficiency of the system it is important that the losses from the thermal storage are as low as possible. For the heat to travel to the storage, heat exchangers are needed, but which specific type will depend of the kind of storage and storage medium used. For CAES a favorable configuration of a TES system is having one hot storage and one cold, albeit interconnected. The cold fluid will flow pass the compressor, cooling it and heating the fluid, which is then stored and later on will flow pass the expander, heating the air and turning the fluid cold yet again (Huang et al., 2018).

16

3. METHODOLOGY

3.1. CASE STUDY A hypothetical scenario is created where 10 average families in Sweden are interested in using a CAES system to cover the peaks in their electricity load profile. The motivation is to save money on electricity cost and to become more independent from the electricity grid with its varying electricity prices. The families all live in the same building and will use one single CAES plant. The plant should be designed to supply electricity to the families during hours of high demand and charge by using electricity from the grid during low demand, all in accordance to the families’ total load profile. The design should be a balance between covering the families’ entire electricity demand during high demand and an economic and physically possible plant design. The design involves establishing the setup of the CAES system, dimensioning the components and determining their characteristics such as pressure ratios etc. The CAES system should be economically evaluated by investigating if there is any arbitrage savings or possibility of revenue being made of selling electricity back to the grid and also considering the cost of the initial investment.

3.2. LOAD PROFILE In order to dimension the CAES system the first thing investigated is the demand that the system must be able to satisfy. The load profile is developed for an arbitrary family of four which is multiplied with 10 to simulate an apartment building. Note that the CAES system is only be used to derive electrical energy (not for heating/cooling purposes). After finalizing the load profile, using data mainly from the energy company (E.ON, 2007) the periods of most intense energy use are identified and so the periods of charge and discharge are decided.

The model is divided into time-‐steps of one hour and spans over one week. For weekdays the energy intensive periods are estimated from 6AM to 9AM in the morning and from 16PM to 22PM in the evening. For weekends the energy intensive period is estimated from 8AM to 22PM. The shortest consecutive time span between the energy intensive periods is seven hours, therefore the compression period is said to be constant of seven hours at all charge periods. To avoid self-‐discharge the end of the charge period is always at the start of a discharge period. During periods of lower energy usage, the CAES system is not providing any energy as the price of electricity generally is lower during these periods. If there is a period where all the mass in not extracted during peak hours, potential surplus energy will be sold during this period.

3.3. CALCULATIONS FOR SYSTEM DIMENSIONING The calculations for system dimensioning are performed from expansion backwards toward compression, since the goal of the system is to supply the energy demand of the apartment building. The first step is calculating how much mass needs to be expanded to provide the required amount of energy. For this, a number of assumptions must be made.

• The compression/expansion can be approximated with isentropic expansion and corresponding coefficient

• The ideal gas law is applicable • Specific heat capacities are constant for each step of the process

The total amount of energy that can be generated by the expansion is:

17

𝐸= = 𝜂?𝜂= ∫ �̇�A𝑤A𝑑𝑡CD (1)

For an open system the work generated by expansion comes from pressure difference and for two-‐stage expansion the work can be divided into two parts, one for each pressure decrease.

𝑤A = 𝑤E +𝑤G = −∫ 𝑣 𝑑𝑝IJIK

− ∫ 𝑣 𝑑𝑝ILMNIJ

(2)

Using the assumption of isentropic expansion where 𝑃𝑣P = 𝑐 is constant and that the ideal gas law is applicable the work for the first expansion can be expressed as:

𝑤E = R 𝑣 𝑑𝑝IK

IJ= 𝑐

EP R

𝑑𝑝

𝑝EP

IK

IJ=

𝜅𝜅 − 1 S𝑝T

𝑐𝑝U

EPV IJIK =

𝜅𝜅 − 1

(𝑝E𝑣E − 𝑝G𝑣G) =

YZY[YZ\Y[]YZ^

𝑝E𝑣E _1 −IJ`JIK`K

a = 𝑐I𝑇E b1 − (IJIK)cdKc e (3)

Applying the exact same theory for the second expansion gives the total work:

𝑤A = 𝑐IE𝑇E b1 − (IJIK)cdKc e + 𝑐IG𝑇G b1 − (

ILMNIJ)cdKc e (4)

Inserting equation (4) into equation (1) gives:

𝐸= = 𝜂?𝜂= ∫ �̇�A(𝑐IE𝑇E b1 − (IJIK)cdKc e + 𝑐IG𝑇G b1 − (

ILMNIJ)cdKc e) 𝑑𝑡C

D (5)

As the air is to be stored in a tank with fixed volume the pressure inside the tank will vary in proportion to the mass inserted or extracted. For each time step the difference in mass inside the tank can be expressed as:

∆𝑚 = ghIhJiAhJ

− ghIhKiAhK

= ghiA(𝑝jG − 𝑝jE) (6)

The total work for one time step, denoting equation (4) as 𝛽 and applying equation (6) to pressure changes during that one time step, then becomes

𝐸= = 𝜂?𝜂=𝛽∆𝑚 (7)

This expression shows that if the demand and turbine and compressor properties are known, the mass needed to meet this demand can be determined.

The process can be divided into several stages where the ideal gas law along with isentropic pressure-‐temperature equation, equation (9), are applied to find the connections between temperature, volume and pressure for each stage. The calculations are performed in Microsoft® Excel where for each time step temperature, volume and pressure are given for each stage. As each time step corresponds to a certain mass and air is assumed to be an ideal gas, equation (8) can be used for each stage. A schematic of how the properties are derived is shown in Figure 2.

𝑝𝑉 = 𝑅j𝑚𝑇 (8)

18

Figure 2. Detailed description of equations used for deciding changes in temperature, pressure and volume for all 22 different stages in the CAES process. Isentropic process relation is abbreviated as IPE and ideal gas law as IGL, both used to denote how the value of the concerned property is derived.

Low pressure compressor inlet

•T = 298,15 K•P = 1 bar•V = IGL

Low pressure compressor outlet

•T = IPE•P = 7 bar•V = IGL

1st intercooler inlet

•Same as highpressure turbine outlet

1st intercooler outlet

•T = 298,15 K•P = 7 bar•V = IGL

High pressure compressor inlet

•Same as previous state

High pressure compressor outlet


2nd intercooler inlet


2nd intercooler outlet

•T = 298,15 K•P = 49 bar•V = IGL

Invisibleexpansion inlet


Isenthalpic Invisible expansion outlet

•Same as next stage

Storage inlet

•T = 298,15 K•P = IGL (based on constant storage volume)•V = IGL

New mass flow

Storage outlet

•T = 298,15 K•P = IGL (based on constant storage volume•V = IGL

Throttle inlet


Isenthalpic Throttle outlet

•T = 298,15 K•P = 25 bar•V = IGL

1st interheater inlet

•Same as previous stage

1st interheater outlet

•T = IPE •P = 25 bar•V = IGL

High pressure turbine inlet


High pressure turbine outlet

•T = 298,15 K•P = 5 bar•V = IGL

2nd interheater inlet


2nd interheater outlet


Low pressure turbine inlet


Low pressure turbine outlet

•T = 298,15 K•P = 1 bar•V = IGL

19

3.4. HEAT TREATMENT During compression heat is produced which ideally will be transferred to a heat storage by installing heat exchangers. This heat will later be used for heating during the expansion. Since the storage is characterized by constant volume and varying losses, the storage needs to be filled to a greater pressure than what is later released, i.e. the outlet pressure of the last compressor is greater than the inlet pressure of the first turbine. Since the air is compressed more than it is later expanded the temperature increase during compression is greater than the decrease during expansion and more heat is generated during compression. This is beneficial for the system in a way, as is accommodates for losses of heat during storage and transfer.

The heat that needs to be stored from the compression is the heat needed to lower the temperature from the temperature increase due to compression to the desired temperature to enter the next compressor or storage. The temperature and pressure relation for isentropic process is:

AJAK= (IK

IJ)cdKc (9)

Assuming that the air should be cooled to the inlet temperature the heat flow which needs to be removed is:

�̇� = �̇�𝑐I(𝑇G − 𝑇E) (10)

This heat flow must be equal to the heat the heat exchangers is able to transfer.

�̇� = 𝑈𝐴∆𝑇mn (11)

It is not within the scope of the report to investigate specific components but if it were so the values of overall heat coefficient, area and temperature difference need to adhere to the relationship in equation (11).

3.5. EFFICIENCIES An issue when measuring the efficiency of a CAES system is the combination of energy sources used (Succar and Williams, 2008). In the two existing plants fuel is burned to provide heat during expansion and electricity to power the compressor. These two sources need to be combined for a total energy input which can be used to define the efficiency. In addition to combining heat and electricity they can be expressed separately as heat rate and charging electricity ration. The heat rate is defined as consumed fuel (Joule) per kWh and is mainly affected by whether it exist a heat recuperation system or not. The charging electricity ratio is defined as generator output by compressor motor input and as mainly affected by piping and throttling losses and efficiencies of compressors and expanders (Succar and Williams, 2008).

Returning to the combined performance index of CAES there are several methods. The simplest way is expressing the efficiency, 𝜂, as a ratio between energy generated by turbine and sum of the energy input to compressor and heating energy. This can be misleading as the energy qualities of the electric and thermal energy differs substantially. Typical values of heat rate and charging electricity ratio gives efficiency of 54% (Succar and Williams, 2008).

20

𝜂 = opoqros

(12)

When combining CAES with nuclear plants, fossil fuel power plants, combined heat-‐power plants (CHP), CAES can convert baseload thermal power into peaking power and thereby contributing to the stability of the electrical grid. For these cases it is relevant to use primary energy efficiency, 𝜂𝑃𝐸, which reflects both the CAES system and the grid. This is done by incorporating the thermal efficiency of the baseload plant. Here, typical efficiency values are around 35-‐40% (Succar and Williams, 2008).

𝜂𝑃𝐸 =op

tqu𝑇ros (13)

An efficiency that is widespread when discussing energy storage solutions is the round-‐trip efficiency, 𝜂𝑅𝑇. It is defined in the same way as in equation (12) but the energy used for heating is adjusted with a second efficiency, 𝜂𝑁𝐺, describing the amount of electricity that could have been produced from the same amount of natural gas in stand-‐alone power plant. The round-‐trip has the advantage of dealing only with electricity inputs and outputs, but the slight disadvantage that 𝜂𝑁𝐺 can be chosen is different ways. The adjusted thermal energy can be placed either as an added input, or as a decreased output, described in equation (14). Depending on the approach round-‐trip efficiencies range between 66 to 88% (Succar and Williams, 2008).

𝜂𝑅𝑇,1 =op

oqrosw𝑁𝐺 or 𝜂𝑅𝑇,2 = op−osw𝑁𝐺

oq (14)

As previously it is beneficial for the system is the heat produced during compression is enough to cover losses for transfer and storage to provide enough heat during expansion. This will reduce all of the efficiency equations to the same expression (ratio between input electricity and output electricity).

3.6. CONSTANTS Various constant values are used in the calculation, both universally accepted constants, various component constants and values approximated as constant.

During all applications of the ideal gas law the gas constant has been set to 𝑅 = 8.3143 |}~m �

and for

air the molar mass has been set to 𝑀 = 29 �}~m

making the specific gas constant 𝑅j = 286.7 |��

.

Constant values that have been set are efficiencies for compressors and turbine, isentropic efficiency to compensate for the error of using isentropic expansion. The pressure ratios have been set to be seven for compression and five for expansion and the tank volume to 15 m3. There is also a constant inlet pressure at which the expansion is working at since this will reduce stress on the system and it is also the method used by the already operating CAES plats (Huang et al., 2018, Succar and Williams, 2008). These four values can be varied but as it is not within the scope of the report to perform complicated optimization they were chosen to be set.

As the system is working at various pressures and temperatures the specific heat capacity varies accordingly. The values are extracted using (Peace Software) and where the specific heat capacity spans over a process with different inlet and outlet properties and average between inlet and average

21

is used. Similarly, the heat capacity ratio or isentropic exponent is assumed to be the constant value 𝜅 = 1,4.

Table 1. Efficiencies for various components and processes.

EFFICIENCIES COMPRESSOR TURBINE (MECHANICAL) TURBINE (GENERATOR) ISENTROPIC

VALUE 80% »1 »1 90% SOURCE (Zhang et al.,

2014) (Salvini et al., 2017) (Salvini et al., 2017) (Salvini et

al., 2017)

Table 2. Specific heat capacities used during calculations.

SPECIFIC HEAT CAPACITY TEMPERATURE(S) [K] PRESSURE [BAR] CP (OR AVERAGE) [J/KG×K]

BEFORE LPC 298 1 1007 1ST INTERCOOLER 520à298 7 1027,5 BEFORE HPC 298 7 1017

2ND INTERCOOLER 520à298 49 1070 1ST INTERHEATER 298à472 25 1043 BEFORE HPT 472 25 1037

2ND INTERHEATER 298à472 5 1021 BEFORE LPT 472 5 1028

3.7. ECONOMIC EVALUATION The economic aspects of CAES is evaluated for the specific conditions and size specified by the case and dimensioning. The created model is used to calculate possible savings and revenues that arise from installing the CAES system. The conclusive cost of electricity is then compared to the cost deduced with the same model but without the CAES system to determine if the system is financially beneficial. The result from the comparison is used to examine if the investment cost of the CAES system to find out if it is profitable.

3.8. ELECTRICITY PRICES To calculate arbitrage savings hourly electricity prices are needed. Using statistical data from Nord Pool for the year 2017 the average of each hour for weekdays and weekends is calculated to match each time step. Added to the price is also the current Swedish electricity tax including VAT. Using the electricity prices, the cost of electricity with and without the CAES system is calculated to determine if CAES gives an income in terms of arbitrage or not.

3.9. INVESTMENTS The necessary initial investments are part of the economic evaluation of the CAES system. These investments consist of all the components needed to run the system. The main part of the total capital cost is assumed to include the capital cost of the compressors, turbines, storage tank and TES. The cost of components beyond the main part is neglected due to the assumptions that the cost of the main part far exceeds the cost of the remaining components such as piping etc. To define the total capital cost of the investments, the individual capital costs of the main components is approximated

22

by analyzing a range of published literature which include economical results of different CAES constructions. Note that a similar approach is adopted in some of the reviewed literatures.

The specific capital cost of the components is calculated by using information about the cost and size of the components found in the reviewed literature. Where possible, an average value is calculated by reviewing several published works and the specific capital cost is then multiplied with the size of the ingoing component in the proposed system. The size of the turbine is set to the highest power the turbine need to achieve to satisfy the demand while the size of the compressor is set to the hourly charging rate for the compression. Finally, the size of the TES-‐system is set to the largest amount of energy the system needs to extract or supply.

All the costs have been converted to 2018 SEK by firstly converting the currency in the literature to SEK using an average exchange rate for the year evaluated (acquired using Oanda Average Exchange Rates (Oanda)) and then using the inflation rate of SEK from the evaluated year to 2018 to convert the cost to SEK in 2018 (acquired using Prisomräknaren (SCB)). To get an estimate of the value of the investment a net present value (NPV) is also calculated on the premise that the electricity used to compress air during charging is obtained free of charge. The annual revenue is consequently the total savings in electricity cost and income from selling electricity stored by the energy storage. The NPV is calculated using Microsoft® Excel which uses equation (15) (Microsoft, 2017).

𝑁𝑃𝑉 = ∑ g�m��j�(Er��C�)�

n��E (15)

In equation (15), n is taken to be the life expectancy of the system which is set to 25 years since this is the lowest value found in reviewed literature (Huang et al., 2018). The discount rate is also found in literature to be around 8-‐10% (Eyer and Corey, 2010 cited in Drury et al., 2011, Huang et al., 2018) which is why 10% is chosen to represent a worst-‐case scenario. The NPV is also calculated for successively increasing electricity prices until the value is positive.

4. RESULTS

4.1. SYSTEM PROPERTIES

4.1.1. LOAD PROFILE For the purpose of deciding the demand of the households, the following load profiles are developed and shown in figure 3. Note that the values of the y-‐axes are for one household and are later multiplied by ten to accommodate ten households. For weekdays the system is set to be discharged between 6:00-‐9:00 and 16:00-‐22:00. As the charge period is seven hours is it set to be charged between 9:00-‐16:00 and 23:00-‐6:00. For weekends the need is greater (note that the y-‐axis is twice as long during weekend) and therefore calculations shows that the system can only supply electricity for a few hours. Therefore, discharge hours are set to be 8:00-‐12:00 and 15:00-‐22:00 to be charged between 12:00-‐15:00 as well as 01:00-‐08:00.

23

Figure 3. Load profiles for one household during Monday-‐Friday and Saturday-‐Sunday.

4.1.2. COMPRESSION AND EXPANSION WORK During the charging phase the two compressors, both working with pressure ratios of seven, using equation (4) demand 449 KJ to compress one kilogram air. Here it is assumed that there are no pressure losses between the two compressors. During the discharge phase the two expanders, working with pressure ratios of five, again using equation (4) provide 359 KJ per kilogram expanded air. The ratio between these two is 80% i.e. for the ideal case of no losses it is only possible to extract 80% of the input work using these pressure ratios.

4.1.3. CHARGE AND DISCHARGE Due to the chosen pressure ratios the storage functions between a maximum pressure of 49 bar (high pressure compressor outlet) and a minimum pressure of 25 bar (high pressure turbine inlet). As it is assumed that the ambient temperature cancels out any temperature variation from pressure changes within the storage the ideal gas law is used to decide how much mass these two pressures represent. The volume of the storage is set to 15 m3. The pressure variation between 49 Bar to 25 Bar corresponds to a mass variation between 860 kg to 440 kg. Figure 4 and figure 5 show that there are periods where the system cannot be used as there is no more mass that can be extracted from the storage. This is clarified in figure 6 where it can be observed that for certain hours that maximum amount of mass that the storage can provide during an entire charge period is not enough for one single hour.

0

500

1000

1500

2000

2500

3000

350000:00

02:00

04:00

06:00

08:00

10:00

12:00

14:00

16:00

18:00

20:00

22:00

Electricity

use [Wh]

Hour of the day

LOAD PROF I LE FOR ONE HOUSEHOLD -‐ WEEKDAY

0

1000

2000

3000

4000

5000

6000

00:00

02:00

04:00

06:00

08:00

10:00

12:00

14:00

16:00

18:00

20:00

22:00

Electricity

use [Wh]

Hour of the day

LOAD PROF I LE FOR ONE HOUSEHOLD -‐ WEEKEND

24

Figure 4. The blue and orange staples show variations in pressure during charge phase, discharge phase and the grey lines follows the pressure in the storage. During periods where only the grey line is present the storage has been emptied to

minimum pressure and the system is at rest. This figure is true for Monday to Friday.

Figure 5. The blue and orange staples show variations in pressure during charge phase, discharge phase and the grey lines follows the pressure in the storage. During periods where only the grey line is present the storage has been emptied to

minimum pressure and the system is at rest. This figure is true for Saturday and Sunday except for the small detail that for the last hour of Sunday the pressure increases like in figure 4 to accommodate the seven-‐hour charge period for Monday.

0,00E+00

1,00E+06

2,00E+06

3,00E+06

4,00E+06

5,00E+06

6,00E+06

00:00

01:00

02:00

03:00

04:00

05:00

06:00

07:00

08:00

09:00

10:00

11:00

12:00

13:00

14:00

15:00

16:00

17:00

18:00

19:00

20:00

21:00

22:00

23:00

Pressure in storage at end

of h

our [Pa

]

Hour of the day

PRESSURE VAR IAT IONS -‐ WEEKDAY

Charge Discharge Storage

0,00E+00

1,00E+06

2,00E+06

3,00E+06

4,00E+06

5,00E+06

6,00E+06

00:00

01:00

02:00

03:00

04:00

05:00

06:00

07:00

08:00

09:00

10:00

11:00

12:00

13:00

14:00

15:00

16:00

17:00

18:00

19:00

20:00

21:00

22:00

23:00

Pressure in storage [Pa]

Hour of the day

PRESSURE VAR IAT IONS -‐ WEEKEND

Charge Discharge Storage

25

Figure 6. The blue and orange staples show the mass that needs to be expanded to satisfy the demand of each discharge period. The grey line is the difference between maximum and minimum mass.

4.1.4. EFFICIENCY To describe the performance of the CAES system the two parameters in questions are the total energy demanded to power the compression and the total energy that can be derived from the expansion process. For one consecutive week the energy needed to power the compressors is 3.00 MJ and the energy provided by the expanders is 1.7 MJ. Using equation (12) to decide efficiency gives a rate of 47%.

4.1.5. HEAT As the pressure ratios are larger for the compression than the expansion, more heat is created during compression than is needed by the expansion. In order to lower the temperature back to inlet temperature during compression 2.5 MJ needs to be removed per kilogram of air. For the expansion 2.0 MJ per kilogram of air needs to be added, giving a ratio of 78%. This ratio allows for heat transfer/storage losses of 22% without having to add external heat during expansion, which will be assumed to be a realistic value.

4.1.6. LOSS DUE TO CONSTANT VOLUME STORAGE As mentioned in the background the storage can be designed to have constant volume or constant pressure. Due to practical reasons constant volume storage (tank) was chosen for this particular report, but this gives greater losses. The expansion pressure losses occur at two stages during each cycle. First when the air is compressed to the maximum pressure by the high-‐pressure compressor and then expands to the pressure of the storage tank. Secondly during discharge when the pressure in the storage goes from maximum to minimum pressure and is expanded through a throttle vault to enter the high-‐pressure turbine at constant pressure. The losses can be calculated using equation (4) and multiplying by the mass passing through that particular stage. For the charging phase equation (4) is used to calculate how much work is lost going from 49 bar to the pressure increase in the tank. For

0

100

200

300

400

500

600

700

00:00

01:00

02:00

03:00

04:00

05:00

06:00

07:00

08:00

09:00

10:00

11:00

12:00

13:00

14:00

15:00

16:00

17:00

18:00

19:00

20:00

21:00

22:00

23:00

Mass [kg]

Hour of the day

MASS NEEDED TO MEET DEMAND FOR EACH D I SCHARGE PER IOD

Weekday Weekend Extractable mass

26

the discharge phase equation (4) is used to calculate how much work is lost going from the pressure decrease in the tank to 25 bar. These losses are calculated as 30.0 MJ.

4.2. FINANCIAL RESULTS

Figure 7. Average electricity spot price for each hour during weekday and weekday for the of 2017 (Nordpool, 2018a).

To determine whether the system can give any form of revenue in terms of arbitrage the electricity price is used for each time step. Above is a graph of how the electricity price varies for each hour during weekday and weekend. It can be observed that the load profiles correspond well to the electricity price variations but also that during one week there are relatively low variations in relation to the absolute price.

Table 3. Economic result for a week with and without CAES.

COST OF COMPRESSION [SEK] INCOME [SEK] TOTAL ELECTRICITY COST [SEK] WITHOUT CAES 0 0 1250 WITH CAES 580 35 1570

Table 4. Capital cost and characteristics of main components.

COMPONENT SIZE SPECIFIC COST CAPITAL COST [SEK]

COMPRESSOR 9.3 [kW] 2700 [SEK/kW] 49 500 SOURCE (Safaei et al., 2012, Huang et

al., 2018, Drury et al., 2011, Madlener and Latz, 2013)

TURBINE 24.6 [kW] 2400 [SEK/kW] 117 000

SOURCE (Safaei et al., 2012, Huang et al., 2018, Drury et al., 2011, Madlener and Latz, 2013)

0

50

100

150

200

250

300

350

400

0 0-‐ 0

1

0 1-‐ 0

2

0 2-‐ 0

3

0 3-‐ 0

4

0 4-‐ 0

5

0 5-‐ 0

6

0 6-‐ 0

7

0 7-‐ 0

8

0 8-‐ 0

9

0 9-‐ 1

0

1 0-‐ 1

1

1 1-‐ 1

2

1 2-‐ 1

3

1 3-‐ 1

4

1 4-‐ 1

5

1 5-‐ 1

6

1 6-‐ 1

7

1 7-‐ 1

8

1 8-‐ 1

9

1 9-‐ 2

0

2 0-‐ 2

1

2 1-‐ 2

2

2 2-‐ 2

3

2 3-‐ 0

0

SEK/MWH

E LECTRICITY PRICE AVERAGES ON WEEKDAY VS. WEEKEND 2017

Weekday Weekend

27

STORAGE TANK 15 [m3] 1750 [SEK/m3] 26 000

SOURCE (Liu et al., 2014) TES 390 [MJ] * 72.5 [SEK/MJ] ** 28 500 SOURCE (Huang et al., 2018)

TOTAL 221 000 * refers to the largest amount of energy needed to be extracted during charging ** calculated by using information about energy needed during a charging

Table 5. Net present value for two scenarios.

SCENARIO CONDITION NPV

1 Zero cost charging electricity -‐79 000 2 Zero cost charging electricity and Higher electricity

price 5 500

It can be observed from table 3 that the electricity cost will actually increase after installing the CAES system. The total electricity cost for the observed week is over 25% higher with CAES compared to without. The reasons behind this will be discussed in later section. Following this conclusion is the fact that the investment won’t be profitable without even considering the initial investment cost of 220 745 SEK that is shown in table 4. Consequently this investment will never yield any return making an NPV-‐calculation obsolete. However, table 5 shows the NPV calculations for two other scenarios. The scenario shows the NPV if the electricity used to compress the air could be obtained free of cost, but this value is negative. The second scenario includes the conditions of the first with the addition of a higher electricity price. The price that yields a positive NPV is approximately 120% larger than the current price.

5. DISCUSSION

5.1. ASSUMPTIONS Various assumptions were made in the model, some more generally used than others. The approximation of isentropic process is adjusted for with isentropic efficiency, which is deemed sufficiently accurate. For the specific heat capacities, the upper and lower value used do not vary with more than 6% so even if an average had been used it would have been quite accurate. Efficiencies for compressors and expanders are approximated according to the literature review and can also be subject to significant uncertainties as they are described to vary within a span of 20 percentage points. Another source of error is the use of the ideal gas law as this is not always the case for air. In conclusion, the isentropic efficiency, specific heat capacity approximations, compressor/turbine efficiencies and ideal gas law all each have a level of uncertainty which should be considered when reviewing the results.

The roughest approximation made is that the wall temperature of the storage tank remains constant. If the tank was perfectly isolated the temperature would vary significantly going back and forth between 25 bar to 49 bar. To calculate the actual temperature variation in the storage, the conducted and convected heat need to be mapped for each time step. Information is needed about the material

28

of the surroundings, the temperature variations of the ground, various heat coefficients, just to name a few parameters. Due to the time and resource limitations of this project this was deemed too complicated to investigate. In the end there were two options for temperature variation in the storage: assume that the ambient cancels out the temperature changes due to pressure variation or assume perfectly isolated walls. Of these two the former was deemed more realistic.

Assumptions were also made in the economical calculations. The capital cost of the components is calculated from economic estimates in previous published works on the subject. The systems in the literature come in a large range of sizes and the costs of these are converted to a specific capital cost depending on the size of the component. The assumption is consequently that economies of scale will not have an impact on the price, which is inaccurate since components of different size do not have the same cost per unit output, but due to the complexity of setting a correct sizing factor the assumption is still made. Costs of labor, maintenance, and components necessary for installation have been neglected. This is due to lack of information and the assumptions that the cost of the considered components will be much larger than these costs which will place these costs within the already large margin of errors in the economical calculations. In summary, with the available information an investment calculation that represent the best-‐case scenario is established which is important to have in mind. In future studies, using a sizing factor to scale up the cost should be considered to more accurately assess the investment cost, if not more accurate information about the cost of specific components can be attained. The uncertainty of the economic evaluation makes these results the one with highest probable error and should therefore be seen as guiding values rather than exact values.

5.2. CONSTANT VOLUME/VARYING PRESSURE VS. CONSTANT PRESSURE/VARYING VOLUME The efficiency for a constant pressure, varying volume system is higher as the compression work and expansion work would be the same if the system is adiabatic. The only losses would be efficiencies of various components and losses from heat transfer. However, if work is the same for compression and expansion heat would have to be added to expansion to keep ambient temperature unless the system is perfectly isolated. Using the same efficiencies for isentropic process, components etc. the round-‐trip efficiency of such a system would be 70%, a significant increase from 47% (note however that TES efficiency needs to be added). However, there are practical difficulties that would need to be met, such as how to create a storage that varies in volume. A solution already mentioned is underwater storage, another solution could be artificial storage with some sort of piston, but this is far more advanced technology.

5.3. OPTIMIZATION There are several values that have been set throughout the CAES process, starting with all the set temperatures. When there is a temperature change the heat exchangers are set to cancel this change out but the system could perhaps be made more efficiency if other temperatures are set. Examples of temperatures that need to be optimized are whether it is truly necessary to lower the temperature to ambient after the compressors. For the expansion side of the process it could perhaps be more efficient to further increase the temperature at the inlets to be able to extract more work. However, as the economic results are subjected to a large error margin and the price of the components are necessary to optimize these parameters it has been judged to be too time demanding for the report.

As was already mentioned in the methodology the pressure ratio and pressure limitations of the storage can be varied to maximize the performance of the system. Higher pressure ratios of

29

compressors and expanders are larger and need a higher power rating which in turn makes them more expensive, so there is a trade-‐off between performance and cost. Working between a larger pressure span within the storage will increase the amount of air that can be extracted, but also increase losses. Thus, there is another trade-‐off between high capacity and round-‐trip efficiency, at least for constant volume storage. Another way to increase the energy storage capacity would be to increase the size of the tank. The size is however limited both by whether there is enough space to accommodate it and also additional cost of larger tanks.

Regarding pressure ratios for both compressor and turbine something that could solve the constant volume related pressure losses would be varying them. However, as this requires more advanced technology, it was decided that constant pressure ratios are more realistic for smaller systems. This is how the already functioning plants are operating, and also gives advantages in terms of reducing stress with a constant power outlet. Another option regarding the compressor and turbine is a combined machine which performs both compression and expansion. This would not change the efficiency, but it could lower the capital cost of the machines since only one machine would have to be acquired instead of two. In conclusion that system could with all likeliness be optimized to improve performance, but this would require detailed information on components and costs, which are unfortunately not available.

5.4. PROFITABILITY One of the possible advantages of energy storage is that it creates a situation where arbitrage is present. The system could use the fact that electricity prices varies according to supply and demand and buy low priced electricity to sell it at a higher price. The results show that this is not the case for this system as the electricity purchased to power the compressor is greater than what is saved by buying electricity at a lower price. This means that even if the capital costs did not exist the system would still cost money.

This leads to the conclusion that if the electricity used to power the compressors needs to be bought from the grid there is no profit whatsoever from using the CAES system. There are two options to make the system profitable in terms of arbitrage: if that energy is produced by the consumer or if that energy is provided free of charge from the grid. If, for example, solar panels are used and produce a surplus that would otherwise not be used or be sold to the grid this energy could be used to charge the storage. As the profits for selling energy to the grid are so low it would be a better option to power the CAES system, despite of the 47% round-‐trip efficiency. The other option would be if the electricity providers would provide the electricity to power the system free of charge. The reason that they would do this is that energy storage reduces the stress on the system for supplying electricity during high demand periods.

An underlying problem for the profitability of CAES is that the electricity prices are so low that even if the CAES can give a large percentage savings of electricity bought from the grid, the price is still too low for the savings to be significant in absolute terms. If the electricity prices rise or the vary more between peak and off-‐peak hours the profitability of a CAES system would improve in terms of arbitrage.

30

5.5. LARGE SCALE OR SMALL SCALE In this report small scale CAES has been investigated, but the existing solutions are all large scale. As mentioned in the background the advantages of CAES can be peak shaving, load levelling, frequency regulation, back-‐up energy provision and possibly arbitrage. Out of these the only one who has the potential of directly affecting the end consumer is arbitrage and as it has already been established this is not present in the current system. It can therefore be concluded that CAES is of interest to the grid as a whole and not the end consumer. This, in combination with the disadvantage of large capital costs, points to large scale CAES being more economically efficient and is likely the reason that the only existing plants are large scale. However, it is possible that smaller CAES plants also have a positive impact on the net as a whole as could help avoid bottlenecks and help bring the supply closer to the user.

5.6. ENVIRONMENTAL ASPECTS CAES has the potential of giving a large positive environmental impact as it could help stabilize the supply of renewable energy. If the energy used to power the compressor comes from a surplus of renewable energy, using CAES would make sure that the energy still is used. On long term this could further integrate renewable energy into the energy system. Much research has been done on implementing CAES with wind power, and one measure that could be taken is to directly connect the CAES system to the wind turbines, thereby eliminating the need to first produce electrical energy to be used for compression. This leads into a more general observation that improving the round-‐trip efficiency would also of course decrease the environmental impact as less energy is needed to produce the same amount of electricity.

Further implementing energy storage solutions into the grid could also eliminate the need for burning coal to supply energy during peak load periods. Just as the main cost of a CAES plant is capital costs, the main environmental impacts correspond to the creating of a storage and the procuring of all components and material used to build the system. To maximize the environmental potential of the CAES plant it is therefore important that the system is built in a resource-‐efficient and sustainable manner, and that the energy later used during the systems life-‐cycle is clean energy.

5.7. SOCIAL ASPECTS Of the so called “three pillars of sustainability” -‐ environmental, economic and social – the smallest overall effect is likely in the social sphere. While its potential for positive environmental effects can result in less fossil fuels being used and thereby having indirect effects on human health and wellbeing, the direct social effects are not as prominent. It could however provide further possibilities to obtain energy security for isolated communities with none or reduced access to the grid. An example of this is remote islands combining wave and/or solar power generation with energy storage. However, isolated communities may not have an excess of available capital to invest in a CAES facility. Another aspect not yet considered in this report is noise pollution, which could have a significant effect especially since the facility considered in this report is meant to be located in close proximity to residential areas.

31

6. CONCLUSIONS AND RECOMMENDATIONS There are several different ways of designing a CAES system and there are elements in the entire cycle from compression to expansion that are of interest from an engineering perspective. This report proposes one system set-‐up alternative but for future studies an important objective is to develop an optimization model that could vary certain properties within the system to determine optimal values to improve efficiency and economic gains. The results from this report suggests that using small scale CAES is not economically viable if the electricity used for charging is bought at today’s electricity price and without any other economic benefits. It could however be made feasible during other circumstances which might inspire further studies on the subject especially since CAES have potential to be a more sustainable alternative to fossil fuels in the energy system. The results show that the positive effects of CAES concern the grid as a whole proposing that the stakeholders with most interest in CAES should be the energy companies and grid developers, not the end consumer. For developing CAES infrastructure it is therefore recommended that this is carried out by the main benefitting stakeholders, rather than the consumers or other smaller actors.

32

7. REFERENCES ABB. 2016. Frequency regulation [Online]. Available: http://new.abb.com/substations/energy-‐storage-‐

applications/frequency-‐regulation [Accessed 2018-‐04-‐10]. BULLOUGH, C., GATZEN, C., JAKIEL, C., KOLLER, M., NOWI, A. & ZUNFT, S. 2004. Advanced Adiabatic

Compressed Air Energy Storage for the Integration of Wind Energy. London UK: Proceedings of the European Wind Energy Conference EWEC 2004.

BYMAN, K. 2016. Electricity production in Sweden. IVA’s Electricity Crossroads project. Royal Swedish Academy of Engineering Sciences, IVA.

CHO, J., JEONG, S. & KIM, Y. 2015. Commercial and research battery technologies for electrical energy storage applications. Progress in Energy and Combustion Science, 48, 84-‐101.

DENHOLM, P., ELA, E., KIRBY, B. & MILLIGAN, M. 2010. The Role of Energy Storage with Renewable Electricity Generation. National Renewable Energy Laboratory.

DONADEI, S. & SCHNEIDER, G. S. 2016. Compressed Air Energy Storage in Underground Formations. DRURY, E., DENHOLM, P. & SIOSHANSI, R. 2011. The value of compressed air energy storage in energy

and reserve markets. Energy, 36, 4959-‐4973. E.ON. 2007. Lilla Energisparboken [Online]. Available:

https://www.ystad.se/globalassets/dokument/lou/avd-‐f-‐strat-‐miljoarb/lilla-‐energisparboken.pdf [Accessed 2018-‐02-‐20].

GARVEY, S. D. & PIMM, A. 2016. Chapter 5 -‐ Compressed Air Energy Storage, Elsevier Inc. HARTMANN, N., VOHRINGER, O., KRUCK, C. & ELTROP, L. 2012. Simulation and analysis of different

adiabatic Compressed Air Energy Storage plant configurations. Applied Energy, 93, 541-‐548. HUANG, Y., KEATLEY, P., CHEN, H. S., ZHANG, X. J., ROLFE, A. & HEWITT, N. J. 2018. Techno-‐economic

study of compressed air energy storage systems for the grid integration of wind power. International Journal of Energy Research, 42, 559-‐569.

IBRAHIM, H., ILINCA, A. & PERRON, J. 2008. Energy storage systems -‐ Characteristics and comparisons. Renewable & Sustainable Energy Reviews, 12, 1221-‐1250.

IPCC (INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE) 2016. Renewable Energy Sources and Climate Change Mitigation. 167.

KUHLMAN, T. & FARRINGTON, J. 2010. What is Sustainability? Sustainability, 2, 3436. LIU, J. C., ZHANG, X. J., XU, Y. J., CHEN, Z. Y., CHEN, H. S. & TAN, C. Q. 2014. Economic Analysis of using

Above Ground Gas Storage Devices for Compressed Air Energy Storage System. Journal of Thermal Science, 23, 535-‐543.

LUND, H. & SALGI, G. 2009. The role of compressed air energy storage (CAES) in future sustainable energy systems. Energy Conversion and Management, 50, 1172-‐1179.

MADLENER, R. & LATZ, J. 2013. Economics of centralized and decentralized compressed air energy storage for enhanced grid integration of wind power. Applied Energy, 101, 299-‐309.

MICROSOFT. 2017. NPV function [Online]. Available: https://support.office.com/en-‐us/article/npv-‐function-‐8672cb67-‐2576-‐4d07-‐b67b-‐ac28acf2a568 [Accessed 2018-‐04-‐26].

MOLINA, M. G. 2017. Energy Storage and Power Electronics Technologies: A Strong Combination to Empower the Transformation to the Smart Grid. Proceedings of the Ieee, 105, 2191-‐2219.

NORDPOOL. 2018a. Historical Market Data: Consumption SE Areas 2017 Hourly [Online]. Available: https://www.nordpoolgroup.com/historical-‐market-‐data/ [Accessed 2018-‐03-‐15].

NORDPOOL. 2018b. The power market [Online]. Available: https://www.nordpoolgroup.com/the-‐power-‐market/ [Accessed 2018-‐04-‐30].

OANDA. Available: https://www.oanda.com/currency/average [Accessed 2018-‐05-‐01]. PEACE SOFTWARE. Available: http://www.peacesoftware.de/einigewerte/luft_e.html [Accessed 2018-‐

05-‐01]. PIMM, A. & GARVEY, S. D. 2016. Chapter 7 -‐ Underwater Compressed Air Energy Storage, Elsevier Inc. SAFAEI, H., KEITH, D. W. & HUGO, R. J. 2012. Compressed air energy storage (CAES) with compressors

distributed at heat loads to enable waste heat utilization. Applied Energy.

33

SALVINI, C., MARIOTTI, P. & GIOVANNELLI, A. 2017. Compression and Air Storage Systems for Small Size CAES Plants: Design and Off-‐design Analysis. In: CAETANO, N. D., FELGUEIRAS, M. C. & FORMENT, M. A. (eds.) 3rd International Conference on Energy and Environment Research, Iceer 2016.

SCB. Available: https://www.scb.se/hitta-‐statistik/sverige-‐i-‐siffror/prisomraknaren/ [Accessed 2018-‐05-‐01].

SHENG, L., ZHOU, Z., CHARPENTIER, J. F. & BENBOUZID, M. E. H. 2017. Stand-‐alone island daily power management using a tidal turbine farm and an ocean compressed air energy storage system. Renewable Energy, 103, 286-‐294.

STOCKHOLM RESILIENCE CENTRE. 2018. The nine planetary boundaries [Online]. Available: http://www.stockholmresilience.org/research/planetary-‐boundaries/planetary-‐boundaries/about-‐the-‐research/the-‐nine-‐planetary-‐boundaries.html [Accessed 2018-‐02-‐09].

SUCCAR, S. & WILLIAMS, R. H. 2008. Compressed Air Energy Storage: Theory, Resources, And Applications For Wind Power. Princeton Environmental Institute, Princeton University.

UN (UNITED NATIONS) 1987. Our Common Future -‐ Brundtland Report. 204. UN (UNITED NATIONS) 2016. The Sustainable Development Goals Report 2016. 24. ZHANG, Y., YANG, K., LI, X. & XU, J. 2014. Thermodynamic analysis of energy conversion and transfer in

hybrid system consisting of wind turbine and advanced adiabatic compressed air energy storage. Energy, 77, 460-‐477.

ZHAO, P., GAO, L., WANG, J. & DAI, Y. 2016a. Energy efficiency analysis and off-‐design analysis of two different discharge modes for compressed air energy storage system using axial turbines. Renewable Energy, 85, 1164-‐1177.

ZHAO, P., GAO, L., WANG, J. F. & DAI, Y. P. 2016b. Energy efficiency analysis and off-‐design analysis of two different discharge modes for compressed air energy storage system using axial turbines. Renewable Energy, 85, 1164-‐1177.

TRITA TRITA-ABE-MBT-18172

www.kth.se

compressed air energy storage - diva1214933/...degree project in technology, first cycle, 15 credits...

Documents