Welcome to the first lesson of the EME 812. In this lesson, we will overview the main types and principles of solar energy conversion to usable outputs, such as electricity, heat, and fuel. There are quite a few technologies that help to do that. Some of those technologies are quite old and well-known, and some are still subject to current research. We will read a couple of recent review papers to learn about those technologies and their impact. Also, we will spend some time reviewing the concept of efficiency, which is a key metric of any process of energy conversion. Also, at the end of this lesson, I will ask you to refresh your knowledge of units and main terms used in the solar energy studies. Let's get started!

Learning Objectives

By the end of this lesson, you should be able to:

1. Explain the basics of solar energy conversion;
2. List examples and parameters of solar systems across the scale;
3. Calculate efficiency of a solar system, based on system performance information.

The energy that is naturally available from the Sun is quite enormous. The Sun delivers 1.2 x 10⁵ $$ TW of radiative power onto the Earth, the amount that surpasses any other energy resource by capacity and availability. That would convert to 3.78 x 10¹² TJ of energy per year. For comparison, according to Crabtree and Lewis (2007), all recoverable Earth's oil reserves (~3 trillion barrels) account for 1.7 x 10¹⁰ TJ of energy. Thus, the sun supplies this amount of energy to the Earth in only ~1.6 days!

A few more stats:

According to reviews of University of Oxford, the current global energy utilization is close to 1.6 x 10⁵  TWh per year (i.e. 5.76 x 10⁸ TJ/year). If we again compare this amount to the global solar energy flux, the Sun is able to cover this demand in only 1 hour and 20 min! This is sort of mind blowing.. 

However, to be utilized, the solar radiation needs to be converted into other forms of energy, such as electricity or usable heat. The question is: can we effectively do that at the scale of our demands?

Evidently, the solar resource contains enough energy to cover those demands. However, the critical limitations in solar energy conversion will be the efficiency of existing technologies and availability of earth materials to scale up those conversion devices.

What's in solar spectrum?

Before considering various types of conversion of solar energy, let us briefly review what solar radiation actually is. Here are a few main things we know from physics:

• Solar energy is electromagnetic radiation.
• Main components of solar radiation reaching the Earth (wavelength, λ, range give in parenthesis):
  • Infrared (52 - 55% $\lambda >\; 700\; nm)$
  • Visible (42 - 43% $400 < \lambda < 700 \mathrm{ n} \mathrm{m }$)
  • Ultraviolet ( 3 - 5% $100 < \lambda < \mathrm{400\; nm}$)
• The most solar radiation reaching the earth surface is essentially in the wave length range $\lambda 290 - 2500$ nm.
• Quantum (unit energy) of electromagnetic radiation - photon (E = hv) - is often a more convenient term in the mechanism of solar conversion.

Figure 1.1. Electromagnetic spectrum showing different types of radiation versus wavelength and frequency scales. The peach box highlights the range represented in downwelling sunlight. 

Credit: Victor Blacus via Electromagnetic Spectrum from Wikimedia Commons

Main types of sunlight conversion

This mix of various types of electromagnetic radiation allows the sunlight to be converted through a variety of physical mechanisms, which are:

• conversion to electricity (photovoltaic effect);
• conversion to usable heat (for example, via thermal collectors);
• conversion to matter / fuel (for example, production of biomass through photosynthesis).

Now we are going to take a closer look at various technologies that are able to convert solar radiation and learn what the main objectives and challenges are there.

Read the following article to overview the main types of solar energy conversion, and try to find the answers to the self-check questions below.

As we perceive from this reading, numerous technologies and areas of research and innovation in solar energy conversion target the overarching objective to raise the device efficiency, thus making it more economically viable for implementation. This is especially true in the light of quite high capital costs for solar energy systems. This challenge is related to both initial materials and manufacturing.

We will talk more about efficiency on the next page of this lesson.

Efficiency is a very important metric in energy conversion. It is most commonly used for evaluating and comparing various methods and devices in terms of technical performance, which is, in turn, related to cost of the technology. The efficiency concept is frequently used in cost estimates and commercial decision making. So, we should spend some time refreshing our basic understanding of the efficiency as a universal metric of conversion systems.

Based on this reading, can you answer the following questions?

We see that efficiency of conversion,η, is a key metric of system performance. When applied to solar energy conversion systems, efficiency of solar energy conversion would be defined as the ratio of the useful output power (delivered by the conversion device) to the incident power (of the solar radiation):

$\eta =\frac{P_{out }}{P_{in }}\times 100\%$

We can answer the following questions from the efficiency analysis:

• What fraction of available energy is lost in the conversion?
• How one device is compared to another?
• What is the performance limit?

PV efficiency measurements

When the efficiency is compared for different types of photovoltaic (PV) cells, we need to make sure that conditions under which the cells are operating are standardized, so that any difference in cell performance is due to the properties of materials and design and not due to the variability of external factors. The nominal efficiency of PV devices is measured at standard conditions [ASTM G173 guide]:

• Air temperature 25°C
• Solar irradiance of 1000 W/m² (clear sky)
• Air mass (AM) of 1.5G
• Cell (panel) oriented perpendicular to the light beam

When the external conditions are kept constant, measured efficiency is solely a device characteristic. To determine efficiency experimentally, we need to measure both the solar irradiance and the power of the cell.

Example of Efficiency Calculation

Generally, to estimate the efficiency of solar energy conversion, you would need:

1. solar irradiance data, and
2. performance data

Consider the example below, which shows estimation of the standard efficiency of a PV module.

Figure 1.2. The connection of efficiency with performance. A solar module of total cell area 2 m²produces a voltage of 45V and a current of 5A at the peak power.

Credit: Mark Fedkin

Standard solar input (irradiance) at the module surface: S = 1000 W/m²

Identifying power input to the PV cell:

$P_{i n }= S = 1000 W / m^2$

Identifying power output from the PV cell:

$P_{ \; \; o \; \; \; u \; \; \; t \; \; } = \; E \; \times \; \frac{I}{ \; \; A \; \; \; r \; \; \; e \; \; \; a \; \; } = \; 112.5 \; \; W \; \; / \; m^2$ (Note: from physics, power is equal voltage times current)

Then, for efficiency, we can write:

$\eta = \frac{P_{o u t }}{P_{i n }}= \frac{112.5}{1000}\times 100 \mathrm{\%} = 11.25 \mathrm{\%}$

Conclusion: only 11.25% of energy flowing to this panel is converted to electricity.

The reason that energy conversion systems have less than 100% efficiency is that there are losses. The origin of those losses can be a complex issue, which could be better understood based on the physics and design of a particular conversion device – PV cell, concentrator, or thermal collector. We will get back to those considerations when talking about specific conversion technologies in detail in respective lessons of this course.

Quality and quantity of solar conversion

There is an important distinction between the total power (measured in Watts) and power density or flux (measured in W/m²). When we talk about the performance of a particular solar energy conversion device (for example, a solar cell), power density characterizes the "quality" of the energy conversion - how much power is generated by each square foot or square meter of the PV cell area. That may depend on properties of the cell material, design, and physical principles behind the conversion process. In contrast, the total power reflects the overall output - the "quantity" of usable energy generated by the whole device per unit of time. In applications of solar energy (say, if we want to power a building), we often look at the total wattage of the system and ways to maximize that total "quantity" of energy supply.

For example, imagine a solar module. At a particular moment of operation, the output power of the device can be expressed as

$P_{o u t }( t o t a l ) = P_{i n }( t o t a l ) \left( \frac{\eta }{100 \mathrm{\%} }\right) = S A \left( \frac{\eta }{100 \mathrm{\%} }\right)$

• η = efficiency (%)
• S = sunlight power density (irradiance) at the cell surface (W/m²)
• A = total cell area (m²)

Logically, to increase the total output from that module, we need to either increase the efficiency or increase the total input power.

The avenue of raising cell efficiency leads us to the physics of the conversion process, material properties, and cell design. The main research and development question here is how to make a better working cell.

The avenue of increasing the total input power leads us to three issues: (i) concentration of light, (ii) sun tracking, and (iii) system scale-up. Concentrating the ambient incident light would indeed increase the amount of energy supplied to the module per unit of time via increasing the S parameter in the above equation. Tracking - i.e., the orientation of the solar panel perpendicular to the sunlight beam - is another way to maximize the amount of absorbable radiation and also contributes to increasing the S parameter. Finally, increasing the size of the module by adding more cells to the system, increasing cell area, or multiplying modules (scale-up) would increase the total active area of conversion (A).

The technology scale-up is the way to match the solar power to commercial applications and consumers' needs. The utility-scale solar power, which is the primary focus of this course, is discussed in the next section.

There are two main solar technologies that are being considered for large scale power generation: (1) Photovoltacs (PV) and (2) Concentrating Solar Power (CSP). Another type - concentrating photovoltaic (CPV) is currently not a major player, but there are a few large facilities that use CPV technology. PV and CSP are principally different in the type of energy conversion and type of solar resource they rely on. We are going to review the basics of those technologies and their current state in the energy market in this lesson before considering more technical details further on.

Photovoltaics (PV)

Video: Energy 101: Solar PV (2:00)

Understanding the limitations in efficiency of solar energy conversion and taking into account the demands of centralized power generation, the technology scale-up is one of the important issues being developed by the government agencies in order to build sustainable energy future.

Obviously, there is a strong push for large-scale systems from the government and industry. But, along with the promise, the scale-up process brings new challenges to the energy conversion system design. Some of those challenges are:

• lower than desired efficiency (theoretical limits suggest it can be much higher);
• high up-front cost of materials and equipment;
• energy storage (electricity or heat);
• power distribution and transmission.

All these issues deserve more attention and will be covered in more detail in further lessons of this course. In this lesson, we are not yet digging into any technical details of the considered technologies but, rather, taking a plunge into the context.

The following materials will give you an idea of the current state of utility scale solar market in the US.

Cencentrating Solar Power (CSP)

The other prominent technology developed on the utility scale in the US and worldwide is Concentrating Solar Power (CSP). While CSP is currently outpaced by PV on the global and domestic market, this technology may be advantageous in the areas with high annual insolation.

Watch this 2-min video to overview the utility-scale Concentrating Solar Power (CSP) systems:

Video: Energy 101: Concentrating Solar Power (2:00)

I hope these materials give you a clear idea what kind of systems will be the subject for learning in this course. The following self-check questions allow you to iterate the basics once more before we move ahead.

Utility-scale solar power installations are on the rise worldwide - the tendency fostered by advances in technology, new energy policies, and markets. Because of this growth, there has been an increased interest among stakeholders to understand the broader impacts of such systems on society and environment. In spite of the often idealistic public perception of solar technology as "green" panacea, an objective examination of the solar technology lifecycle reveals both positive and negative impacts.

Large photovoltaic plants are often perceived as competition for farming due to their substantial land footprint.

Image by Samuel Faber from Pixabay

Careful impact assessment of large solar projects is important in order to steer the energy infrastructure development towards the optimal solutions that would take into account economic, environmental, and social values. Understanding the sensitivities and existing ecosystem services at the locale at the utility project planning stage is becoming a key step in responsible solar development. 

Please read the following review article, which nicely covers the multiple effects of utility solar power.

At the conclusion of this lesson, I want to refer you to some resources on basic energy units, conversions, and terminology that specifically applies to solar energy systems. If you have just completed the EME 810 course, you will find many of these things familiar. For example, can you clearly answer these questions:

• What is the difference between power and energy? And what are their units?
• What is the difference between power and power density?
• What is irradiance? What is irradiation? Is there difference?
• What are the symbols for irradiation on hourly and daily steps?

Video (8:46)

Also, it would be useful to look through the original technical paper by Beckman et al. (1978), too, and use it in the future if any notation questions arise. The main purpose of this material is for everyone to be on the same page when analyzing the solar energy conversion technologies further in this course.

Readings and activities in this lesson give you a general perspective of this course and set the context without yet addressing the specific science of the solar energy conversion technologies. Here, we try to figure out what aspects and what impacts would be important when the conversion technologies are scaled up to the utility level. Hopefully the materials of this lesson also provided you with a good refresher of such basic concepts and terms as energy conversion, efficiency, power, power density.

In Lesson 1, we learned that the main function and purpose of the solar energy systems is to convert sun radiation - i.e., light or heat - into electricity. However, the efficiency of such conversion is not very high. One way to make many known solar technologies feasible with respect to their efficiency, total output, environmental impact, and cost is to concentrate the incoming radiation. Concentration of light will be the main topic of Lesson 2.

Sunlight is a practically inexhaustible natural resource which is also universally available. However, one of the disadvantages or difficulties related to its utilization is a relative low density of the solar flux. To generate sufficient power to meet demands of large populated zones, a vast area should be covered by solar collectors, and a significant amount of materials and resources should be spent on production and service of those collectors. This expense raises a question about economic viability of solar and initiates the search for ways to increase the sunlight conversion efficiency one way or the other. Generally, there are two ways to solve the problem - to improve the conversion device (intrinsic factor) or to increase the input flux (extrinsic factor). While the first avenue is subject to energy engineering research and innovation (e.g., developing new types of photovoltaic materials and devices), the second option - concentration of the incident solar flux - is already widely implemented. This lesson presents basic concepts for sunlight concentration and discusses typical optical geometries common in utility scale solar plants. This material provides background for further discussion of such technologies as concentrating solar power (CSP) or concentrating photovoltaics (CPV) later in this course.

Learning Objectives

By the end of this lesson, you should be able to:

1. understand the physical principles of light concentration;
2. list the main types of concentration systems used for utility scale solar facilities;
3. calculate parameters of the typical light concentrating systems (CPC, parabolic concentrators).

Before talking about concentration of light for practical purposes, it would be good for us to review what kinds of natural radiation are available to us and how that radiation is characterized and measured.

Solar Constant

The fraction of the energy flux emitted by the sun and intercepted by the earth is characterized by the solar constant. The solar constant is defined as essentially the measure of the solar energy flux density perpendicular to the ray direction per unit area per unit of time. It is most precisely measured by satellites outside the earth atmosphere. The solar constant is currently estimated at 1361 W/m² [cited from Kopp and Lean, 2011]. This number actually varies by 3% because the orbit of the earth is elliptical, and the distance from the sun varies over the course of the year. Some small variation of the solar constant is also possible due to changes in Sun's luminosity. This measured value includes all types of radiation, a substantial fraction of which is lost as the light passes through the atmosphere [IPS - Radio and Space Services].

Transformations in the Atmosphere

As the solar radiation passes through the atmosphere, it gets absorbed, scattered, reflected, or transmitted. All these processes result in reduction of the energy flux density. Actually, the solar flux density is reduced by about 30% compared to extraterrestrial radiation flux on a sunny day and is reduced by as much as 90% on a cloudy day. The following main losses should be noted:

• absorbed by particles and molecules in the atmosphere - 10-30%
• reflected and scattered back to space - 2-11%
• scattered to earth (direct radiation becomes diffuse) - 5-26% [Stine and Harrigan, 1986]

As a result, the direct radiation reaching the earth's surface (or a device installed on the earth's surface) never exceeds 83% of the original extraterrestrial energy flux. This radiation that comes directly from the solar disk is defined as beam radiation. The scattered and reflected radiation that is sent to the earth's surface from all directions (reflected from other bodies, molecules, particles, droplets, etc.) is defined as diffuse radiation. The sum of the beam and diffuse components is defined as total (or global) radiation.

It is important for us to differentiate between the beam radiation and diffuse radiation when talking about solar concentration in this lesson, because the beam radiation can be concentrated, while the diffuse radiation, in many cases, cannot. For that matter, the solar systems utilizing concentrating collectors will work best in sunny locations and may not be feasible in those with a lot of weather variability and clouds.

Solar Radiation Metrics

Consider the following metrics commonly used to report the solar resource (irradiance) data. These values can be determined from the field measurements or from empirical correlations.

Theoretically, these three metrics are interrelated:

However, in practice, field measurements may somewhat deviate from this relationship.

A typical solar resource data file (Typical Meteorological Year or TMY) would include all of these metrics measured for a specific location for each hour for each day in a year. Note that these values (measured in W/m²) indicate the instantaneous solar flux, which of course will vary during the day. In the morning and in the evening, the irradiance will be lower, but it will often reach its peak around solar noon. If there are clouds or other weather phenomena, the irradiance will temporarily drop.

The plots below give you an example of such variance. The GHI, DNI, and DHI data are plotted for the day of March 21st (equinox) in Orlando, FL. While it seemed to be a relatively sunny day (the beam component evidently dominates over diffuse, reaching ~900 W/m²), there are some minor interruptions (possibly from clouds) to this profile.

Daily profiles for different types of solar radiation in Orlando, FL, on March 21 - extracted from the TMY weather file. 

Credit: Mark Fedkin

The TMY files with all these metrics given for each day for different locations around the globe are publicly available from the NSRDB database.

Solar Maps

The GHI data are also used to generate solar resource maps. However, the instantaneous values of global irradiance are not best for mapping due to their continuous variability. Instead, GHI are integrated to determine the daily average irradiation (total energy from the sky).

Look again at the GHI plot (blue curve above) – essentially, this total daily energy will be equal to the area under the irradiance curve! This total daily irradiation value (measured in kWh/m²/day) can be better related to the total energy converted and delivered by your solar system. In a practical sense, it is a more intuitive metric to map.

Also, let’s not forget the seasonal variations. The solar daily irradiation will be understandably higher during summer months and lower during winter months. Hence, the map below is based on the annual average values of daily irradiation.

Solar Irradiation Map for the U.S. in terms of annual average daily GHI (kWh/m²/day)

Sengupta, M., Y. Xie, A. Lopez, A. Habte, G. Maclaurin, and J. Shelby. 2018. "The National Solar Radiation Data Base (NSRDB)." Renewable and Sustainable Energy Reviews 89 (June): 51-60.

National Renewable Energy Laboratory

Short-Wave and Long-Wave

Short-wave radiation, in the wavelength range from 0.3 to 3 μm, comes directly from the sun. It includes both beam and diffuse components.

Long-wave radiation, with wavelength 3 μm or longer, originates from the sources at near-ambient temperatures - atmosphere, earth surface, light collectors, other bodies.

The solar radiation reaching the earth is highly variable and depends on the state of the atmosphere at a specific locale. Two atmospheric processes can significantly affect the incident irradiation: scattering and absorption.

Scattering is caused by interaction of the radiation with molecules, water, and dust particles in the air. How much light is scattered depends on the number of particles in the atmosphere, particle size, and the total air mass the radiation comes through.

Absorption occurs upon interaction of the radiation with certain molecules, such as ozone (absorption of short-wave radiation - ultraviolet), water vapor, and carbon dioxide (absorption of long-wave radiation - infrared).

Due to these processes, out of the whole spectrum of solar radiation, only a small portion reaches the earth's surface. Thus, most x-rays and other short-wave radiation is absorbed by atmospheric components in the ionosphere, ultraviolet is absorbed by ozone, and not-so abundant long-wave radiation is absorbed by CO2. As a result, the main wavelength range to be considered for solar applications is from 0.29 to 2.5 μm  [Duffie and Beckman, 2013].

Figure 2.1. Different types of radiation at the earth surface: orange - short wave; blue - long wave.

Credit: Mark Fedkin - modified after Duffie and Beckman, 2013

Transmittance

The effects of radiation scattering and absorption vary with the time of the day (due to the change of the air mass through which the beam passes through) and seasonally with the time of the year. Hence, the actual beam irradiance on the surface can be empirically estimated using a set of atmospheric parameters and Sun-Earth geometry.

Hottel’s method (Hottel 1976) describes the beam radiation transmitted through the atmosphere under the “clear-sky” conditions using atmospheric transmittance coefficient.

$$G_{b n }= {\tau }_b\times G_{o n}$$

where G _bn is the beam irradiance normal to the receiving surface, G _on is the extraterrestrial irradiance (solar constant in a general case), and τ _b is beam transmittance.

The transmittance value can be evaluated by Hottel’s model using solar zenith angle and altitude for several different climate regimes. Or, it can be determined by direct measurement of the beam irradiance on the normal surface.

Instruments

The amount of solar radiation on the earth's surface can be instrumentally measured, and precise measurements are important for providing background solar data for solar energy conversion applications.

Described below are the most important types of instruments to measure solar radiation:

1. Pyrheliometer is used to measure direct beam radiation at normal incidence. There are different types of pyrheliometers. According to Duffie and Beckman (2013), Abbot silver disc pyrheliometer and Angstrom compensation pyrheliometer are important primary standard instruments. Eppley normal incidence pyrheliometer (NIP) is a common instrument used for practical measurements in the US, and Kipp and Zonen actinometer is widely used in Europe. Both of these instruments are calibrated against the primary standard methods.

   Based on their design, the above listed instruments measure the beam radiation coming from the sun and a small portion of the sky around the sun. Based on the experimental studies involving various pyrheliometer design, the contribution of the circumsolar sky to the beam is relatively negligible on a sunny day with clear skies. However, a hazy sky or a uniform thin cloud cover redistributes the radiation so that the contribution of the circumsolar sky to the measurement may become more significant.

2. Pyranometer is used to measure total hemispherical radiation - beam plus diffuse - on a horizontal surface. If shaded, a pyranometer measures diffuse radiation. Most of solar resource data come from pyranometers. The total irradiance (W/m²) measured on a horizontal surface by a pyranometer is expressed as follows:

   $$I_{t o t }= I_{b e a m }\cos \theta + I_{d i f f}$$

   (2.1)
    

   where θ is the zenith angle (i.e., angle between the incident ray and the normal to the horizontal instrument plane.

   Examples of pyranometers are Eppley 180^o or Eppley black-and-white pyranometers in the US and Moll-Gorczynsky pyranometer in Europe. These instruments are usually calibrated against standard pyrheliometers. There are pyranometers with thermocouple detectors and with photovoltaic detectors. The detectors ideally should be independent on the wavelength of the solar spectrum and angle of incidence. Pyranometers are also used to measure solar radiation on inclined surfaces, which is important for estimating input to collectors. Calibration of pyranometers depends on the inclination angle, so experimental data are needed to interpret the measurements.

3. Photoelectric sunshine recorder. The natural solar radiation is notoriously intermittent and varying in intensity. The most potent radiation that creates the highest potential for concentration and conversion is the bright sunshine, which has a large beam component. The duration of the bright sunshine at a locale is measured, for example, by a photoelectric sunshine recorder. The device has two selenium photovoltaic cells, one of which is shaded, and the other is exposed to the available solar radiation. When there is no beam radiation, the signal output from both cells is similar, while in bright sunshine, signal difference between the two cells is maximized. This technique can be used to monitor the bright sunshine hours.

   A more detailed explanation of how these instruments work and what kind of data is obtained from those measurements is available in the following Duffie and Beckman (2013) book, referred below. Please spend some time acquiring basic knowledge on solar resource data. For everyone who took EME 810 and is more or less familiar with this topic, this still may be a useful refresher.

Solar radiation data collected through the above-mentioned instrumental methods provide the basis for development of any solar projects. We can summarize the types of solar resource data as follows:

Figure 2.2. Overview of different types of solar radiation data.

Click here for a text description

Solar radiation data supplied via pyrheliometric and pyranometric measurements may represent time resolved information: e.g., irradiance (instantaneous measurements of solar energy flux), irradiation (integrated energy flux over time), or averaged irradiation. Depending on measurement setup, the data can be for horizontal or inclined surface. The data can characterize different types of radiation: beam, diffuse, or total.

Credit: Mark Fedkin

Before moving on, please work through the following self-check questions to assess your learning:

Any general setup for the conversion of the solar energy includes a receiver - a device that is able to convert the solar radiation into a different kind of energy. This can be either a heat absorber (to harvest thermal energy) or a photovoltaic cell (to convert light to electric energy). In the first case, the thermal radiation is absorbed to heat a medium (fluid), which transfers that absorbed energy to a generator. In the second case, light causes a photovoltaic effect in the material of the solar cell, which generates electric current. In both of these situations, the amount of energy available for the conversion is only as much as the solar source supplies per unit area of the converter.

If we need more energy for use, we have two options. The first option is to increase the system scale (for example by increasing the number of receivers). In other words, we have to expand the plant area, which would involve additional cost for construction, service, maintenance, and may require additional land, more materials, etc. It has been done to some extent, but sometimes it is not a sufficient measure to meet the energy demands, especially if land area is a constraint. The second option is to concentrate the radiation flux. This can be achieved by placing a concentrator (usually some kind of optical device) between the light source (sun) and the receiver. By common terminology, a solar collector is a sunlight processing system that includes a concentrator and a receiver in its setup; it is also characterized by aperture - the cross sectional area through which sunlight accesses the system.

The most common concentrators are reflectors (mirrors) and refractors (lenses), which modify and redirect the incident sunlight beam. The design of the concentrating optics varies. Some of the examples of concentrating collectors, which involve diversely shaped mirrors, are shown in Figure 2.3, as they applied to the solar-to-thermal energy conversion.

Figure 2.3. Types of concentrating sunlight collectors: (a) tubular absorbers with diffuse back reflector, (b) tubular absorbers with specular cusp reflectors, (c) plane receiver with plain reflectors (V-trough), (d) multisectional planar concentrator, (e) compound parabolic concentrator (f) parabolic trough, (g) fresnel concentrator, (h) array reflectors (heliostats) with central receiver. Concentration of light on the receiver is achieved by shaping the reflectors (mirrors) around the receiver (represented by blue circles).

Credit: Mark Fedkin - modified after Duffie and Beckman, 2013

The process of light concentration implies first of all that the energy flux is increased due to confining it to a smaller area. This brings several important benefits:

• reaching higher temperatures for heat collectors;
• heat losses from the surface of the receiver are decreased because the receiving area is decreased;
• higher energy conversion rate can be achieved over a smaller area.

There are two major classes of solar concentrators: imaging and non-imaging. Imaging concentrators are called imaging because they produce an optical image of the sun on the receiver. Non-imaging concentrators do not produce such an image, but rather disperse the light from the sun over the whole area of the receiver. Non-imaging concentrators have relatively low concentration ratio (<10) compared to the imaging concentrators.

All of the optical tools designed for manipulating sunlight for the purpose of its concentration and efficient utilization are based on the fundamental optics principles, which you may remember from physics courses. In case you need to refresh your knowledge of those fundamentals before we study the light concentration principles, please refer to the following reading and video:

Out of the different types of concentrators listed above, mainly the following four technologies have been adopted for use in the utility scale CSP facilities [Mendelsohn et al., 2012]:

1. Parabolic trough
2. Solar tower
3. Parabolic dish
4. Linear Fresnel reflector

All of these are imaging concentrators which allow relatively high concentration temperatures: about 400 ^oC for parabolic troughs, up to 650 ^oC for Stirling dishes, and above 1000 ^oC for solar power towers. Just for comparison, non-imaging concentrators would work maximum up to 200 ^oC. These technologies will be introduced in more detail in Lessons 7 and 8 of this course.

There are also developments for non-imaging compound parabolic collectors (CPC) to be used at the utility scale for low-temperature applications [Baig et al., 2009], but this technology is not as widespread due to its moderate concentrating capabilities. Its flexibility with respect to using non-beam radiation and more relaxed technical requirements to positioning of concentrators are still attractive, so this technology will be also included in our consideration.

Concentrating photovoltaics is another technology class that uses concentrated light, but those devices will be covered separately in Lessons 5 and 6 of this course.

The light concentration process is typically characterized by the concentration ratio (C). By physical meaning, the concentration ratio is the factor by which the incident energy flux (I_o) is optically enhanced on the receiving surface (I_r) - see Figure 2.4. So, confining the available energy coming through a chosen aperture to a smaller area on the receiver, we should be able to increase the flux.

In the above equation, C_geo is called the geometric concentration ratio. It is easy to use, as the areas of the devices are known, although it is adequate only when the radiation flux is uniform over the aperture and over the receiver. Also, please note that for some imaging concentrators, the area of the available receiver surface can be different from the area of the image produced by the concentrator on the receiver. So, if the image does not cover the entire surface of the receiver, we need to use the image area to estimate the concentration ratio.

Figure 2.4. Schematic representation of light concentration process.

Credit: Mark Fedkin

The concentration ratio can also be represented by the energy flux ratio at the aperture and at the receiver. In this case, it is termed optical concentration ratio C_opt (or flux concentration ratio) and can be directly applied to thermal calculations.

In case the ambient energy flux over the aperture (insolation) and over the receiver (irradiance) is uniform, the geometric and optical concentration ratios are equal (C_geo = C_opt).

The concentration ratios are important metrics used to characterize and rank optical concentrators. Next, we will look at several examples of concentrator designs and see what values of concentration ratios they can provide.

There is a theoretical limit to solar concentration. For circular concentrators - 45,000, and for linear concentrators - 212, based on the geometrical considerations; however, these limits may be unreachable by real systems because of non-idealities and losses. If you are interested in the analytical estimation of the concentration limits, refer to Duffie and Beckman's (2013) book (p.325) for more details.

In general sunlight, concentration systems are roughly classified into: low concentration range (C<10), medium concentration range (10<C<100), and high concentration range (C>100). However, only some of the systems provide uniform concentrated light flux (e.g., V-troughs or pyramidal plane reflectors) and can be characterized by a single concentration ratio. Many systems with curved reflecting surfaces (e.g., conical, parabolic, spherical) create a distribution of flux density over the receiver and would rather be characterized by a variable C over the receiver width. In that case, a local concentration ratio (C_l) is the main parameter to characterize the performance of the ideal concentrator:

where I(y) is determined for any local position y from the center of the produced image, and I_ap is the intensity of the incident radiation at the aperture.

Figure 2.5. Example of distribution of local radiation intensity over the receiver of an imaging collector.

Credit: Mark Fedkin after Duffie and Beckman, 2013

In many typical cases of imaging concentrators, the reflectance of the surface (ρ), i.e., the fraction of light radiation reflected from the surface compared to the total incident radiation, is also taken into account. Then the local intensity of the concentrated light, I(y), can be described as follows:

Further, in this lesson, we will study some examples that use this equation to estimate energy distribution within a concentrated image on the receiver. It would be better to have a specific type of concentrator to apply these concepts. Please read through the following text to enforce your understanding of the concentration ratios.

Parabolic geometry is the basis for such concentrating solar power (CSP) technologies as troughs or dishes. Parabolic trough is also considered one of the most mature and most commercially proven technologies in the utility scale CSP facilities (Mendelsohn et al., 2012), so we will look at the physical principles of parabolic concentrators in some more detail.

Geometrically, a parabola is a locus of points that lie on equal distance from a line (directrix) and a point (focus) - see Figure 2.6. For each point of the parabola, DR = FR. The distance VF between the vertex and focus of the parabola is the focal distance (f). The line perpendicular to the directrix that passes through the focus is the axis of the parabola; the axis divides the parabola into two parts that are symmetrical.

With origin at its vertex, and the axis of the parabola taken as x-axis, a parabola is described by the equation:

where f is the focal length.

Figure 2.6. Geometry of the parabola.

Credit: Mark Fedkin

By definition of the focal point of the parabola, all incoming rays parallel to the axis of the parabola are reflected through the focus. This provides an opportunity for light concentration by using parabolic surfaces. If we assume that solar light arrives to the surface as essentially parallel rays, and apply the Snell's law (the angle of reflection equals the angle of incidence), we can assign the focal point as an ideal location for the receiver (Figure 2.7).

Figure 2.7. Geometry of a parabolic reflector. All rays parallel to the parabola axis are reflected through the focal point.

Credit: Mark Fedkin

Solar applications deal with a parabola of a finite height (Figure 2.8). The design of the parabolic reflector takes into account the available aperture size (a), focus location (f - i.e., where receiver would be placed), and height of the reflector (h). These parameters are interrelated via the equation (Stine and Harrigan, 1986):

Figure 2.8. Corresponding variation of the parabolic curvature and focal distance for a fixed aperture.

Credit: Mark Fedkin

This figure above shows that the flatter the reflecting surface, the longer the focal length. The "flatness" of the shape of a finite parabola is typically characterized by the rim angle ($\varphi$). When rim angle increases (within the same aperture), the parabola becomes more curved, and the focal distance shortens.

Figure 2.9. Parabolic trough mirror installation on a solar farm.

Credit: Kuraymat (via Flickr)

Parabolic trough (Figure 2.9) is a typical example of an imaging concentrator that utilizes the geometric relationships discussed above. Parabolic trough is one of the most widely implemented technologies for sunlight concentration at the utility scale. This type of collectors relies on sun tracking to ensure that the beam radiation is directed parallel to the parabolic axis.

A parabolic mirror produces an image of the sun on the surface of the receiver, so the receiver size needs to be matched to the image size. Consider Figure 2.10, which illustrates this idea. Since the sun is not really a point source, solar beam incident on the reflector is represented as a cone with an angular width 0.53^o (so the half-angle between the cone axis and its side is 0.267^o). Being reflected at a point on the parabolic surface, the beam hits the focal plane, where it produces an image of a certain dimension, centered around the focal point. The diameter of the cylindrical receiver (D), which would intercept the entire reflected image can be theoretically calculated using aperture width (a), and rim angle ($\varphi$) as follows (Duffie and Beckman, 2013):

Figure 2.10. Beam reflection in an imaging parabolic concentrator with aperture a. Blue area imitates the position of a cylindrical (tubular) receiver at the focal point. Gray area imitates a flat receiver at the focal plane. The minimal size of the receiver to intercept all of the reflected radiation should be matched to the diameter of the reflected beam at the focal plane.

Credit: Mark Fedkin - modified after Duffie and Beckman, 2013

For the linear receiver, the width of the image (W) produced on the focal plane can be determined as follows:

The equations presented here can be used to estimate the size of the reflected light image on the receiver for different shapes of parabolic reflectors. The formulas include a as a chosen aperture of the reflector (width of the trough), and ($\varphi$) as a measure of parabolic curvature. Note that these are the minimal theoretical dimensions of the reflected image that would be produced by the ideal parabolic mirror that is perfectly aligned. If there are any flaws in the mirror surface or trueness of the angle, additional spreading of the image may occur. If you are interested in more explanation of how these formulas were derived, please refer to Duffie and Beckman, 2013 book (Section 7.9)

The above-described geometrical concepts apply to the cross-section of a parabolic reflector. In reality, the reflector itself is a three-dimensional shape, i.e., a parabolic cylinder with a finite length (l). So, the cone-shaped ray reflected at a point on the surface of a parabolic reflector will produce an ellipse-shaped image on the focal plane. We can see that as the reflection point is moved away from the vertex towards the rim, the ellipse transforms from a circular to a more and more elongated shape (because the cone would be sectioned by the focal plane at greater and greater angle - Figure 2.11).

Figure 2.11. Cross section of a cone (shown as gray shaded area) in the direction perpendicular to its axis (left) and at an angle to its axis (right).

Credit: Mark Fedkin

Knowing the angular width of the cone, the dimensions of the ellipse image can be theoretically derived and presented as a function of $\varphi$ (angle of deviation from the parabola axis). Below are the equations describing the length of the minor and major axes of the ellipse.

where r is the distance between the focus and reflection point (local radius) on the parabolic mirror (r=f at the vertex); $\varphi$ is the angle between the parabola axis and the ray, and 0.267^o is the half-angle of the ray cone width.

The superposition of these individual ellipses produced by each element of the reflector form the total image, which is not uniform, but rather has a distribution of light intensity. The focal length (which is related to the rim angle of the reflector) is responsible for image size, while the aperture is responsible for the total amount of energy concentrated by a collector. So, the total image intensity (brightness) at the receiver should be a function of a/f. The image brightness essentially reflects the energy flux concentration:

The distribution of intensity of the energy flux within the concentrated image may have a profile similar to Figure 2.5. Different models have been applied to quantify that profile. For example, one of the approaches is called nonuniform solar disk, which suggests that the sunlight intensity coming out of the center of the solar disk is higher than that coming from its edges [Evans, 1977]. Without going into too much detail of this model, we can use the diagrams presented in the book by Duffie and Beckman (2013), which allow connecting various parameters of a parabolic concentrator with the local intensity on the receiver.

Please refer to the following reading to study the tools for image analysis via the nonuniform solar disk model, and be sure to study the example presented therein, which is very helpful.

Please answer the following quick questions to check your understanding of some of the basic points in this section. The parabola cheatsheet presents a useful summary for your notes. 

The compound parabolic concentrators (CPC) are typical representatives of non-imaging concentrators, which are capable of collecting all available radiation - both beam and diffuse - and directing it to the receiver. These concentrators do not have such strict requirements for the incidence angle as the parabolic troughs have, which makes them attractive from the point of view of system simplicity and flexibility. Like parabolic and other shapes, CPC concentrators can be applied in both linear (troughs) and three-dimensional (parabolocylinder) versions. The same as in "pure" parabola case, troughs are most widespread and useful for this type of concentrator.

The geometry of a CPC collector is demonstrated in Figure 2.12. If we consider a CPC trough, this diagram represents its cross-section. Each side of the shape is a parabola, and each of the parabolas has its focus at the lower edge of the other parabola (e.g., F is the focus of the right-hand parabola in Figure 2.12). Each parabola axis is tilted relative to the axis of the CPC shape. One of its key parameters is acceptance half-angle (${\theta }_c$), which is the angle between the axis of the collector and the line connecting the focus of one of the parabolas with the opposite edge of the aperture. The collector is designed in such a way that each ray coming into the CPC aperture at an angle smaller that $\theta$ reaches the receiver; if this angle is greater than ${\theta }_c$, the ray will return (Figure 2.13). The relationship between the size of the aperture (2a), the size of the receiver (2a') and the acceptance half-angle is expressed through the following equation:

Knowing that the geometric concentration ratio is the quotient of the aperture area to the receiver area (see Section 2.3), for a linear CPC concentrator, we can obtain the relationship between the concentration ratio and the acceptance angle:

Figure 2.12. Geometry of the compound parabolic concentrator (CPC).

Click for a text description of the image

One large parabolic mirror with a second mirror sitting tangent to the parabolic axis with an end at mirror #1’s focus. The distance between the two upper ends of the parabolas is labeled aperture (2a) and the bottom two ends is labeled receiver. Dashed lines connect one top end to the opposite bottom end. The angle between their y intercept, y-axis and upper tip represents the acceptance half-angle.

Credit: Mark Fedkin after Duffie and Beckman, 2013

Figure 2.13. Reflection of light directed to the CPC concentrator at different angles. All rays with the angle of incidence smaller or equal to the acceptance angle reach the receiver at the bottom; all rays with the angle of incidence higher than the acceptance angle are reflected back to the aperture.

Credit: Mark Fedkin

There are some other useful expressions that describe the design of CPC concentrators. The following equations relate the focal distance of the side parabola (f) to the acceptance angle, receiver size, and height of the collector (Duffie and Beckman, 2013):

Please complete the following reading to further explore the work principle of CPC concentrators.

The following self-check questions will help you to test check your learning of the principles of CPC collectors:

Lesson 2 covers fundamental principles of light concentration that are important for a number of solar energy conversion technologies - both thermal and photovoltaic conversion. The general scheme of the solar energy concentration is this:

Input solar energy flux ⇒ Optical concentration device ⇒ Output concentrated solar energy flux

We touched upon each of these stages. First, we looked at the available solar radiation at the earth surface - the input we start with. Then, we considered a few techniques that concentrate the available flux, confining it to a smaller area. Finally, we looked at the output and its characteristics. Theoretical and empirical laws presented in the readings provide you with the background for estimating such parameters as concentration ratio and output energy density. Most of the theoretical considerations presented here are made for ideal systems. In reality, you can expect that imperfect optics will require additional corrections for non-ideality and losses. Limitations and advantages of specific concentrating technologies will be considered in further lessons, separately for CSP and photovoltaic systems.

After you have covered the assigned materials for this lesson, please complete the following assignments:

This lesson will introduce the concept of sun tracking and will discuss how it can improve the performance of solar energy systems. The sun is a light source that is not fixed but rather is constantly moving relative to a solar receiver. This leads to significant variability of the available radiation and, as a result, variability of power output and efficiency of a solar energy conversion system. The idea of sun tracking was developed in an attempt to mitigate that variability to some extent and in pursuit of higher efficiency and extending the solar power production over the course of the day. Tracking technology is more often associated with utility scale solar plants rather than small residential systems. Some examples of tracking include single-axis and two-axis tracking of PV panels, moving heliostats in solar tower thermal plants, variable tilt parabolic trough systems, and Stirling dish concentrators - systems whose operation heavily relies on the accuracy of tracking. In this lesson, we will first discuss when tracking is a viable idea and what systems can benefit from it. Then, we will study the geometry of the solar motion through the sky and define the parameters that characterize the position of the sun relative to a solar receiver at a certain location and time. This background would be important in understanding any tracking algorithms. Some examples and activities within this lesson will involve geometric calculations that will help you to better understand how this technology works.

Learning Objectives

By the end of this lesson, you should be able to:

1. define the main parameters of the solar motion;
2. explain the types of tracking systems and principles of their operation;
3. calculate the position of the sun relative to the receiving surface at a locale at a particular time.

Solar tracking is a technology for orienting a solar collector, reflector, or photovoltaic panel towards the sun. As the sun moves across the sky, a tracking device makes sure that the solar collector automatically follows and maintains the optimum angle to receive the most of the solar radiation. Some solar concentrators hugely benefit from tracking, while some others do not. So, the tracking systems can be added with additional cost and certain trade-offs in system design only when it pays off.

The required accuracy of tracking varies with application. For example, concentrators, especially in solar cell applications, require a high degree of accuracy to ensure that the concentrated sunlight is directed precisely to the solar conversion element. Tracking the sun from east in the morning to west in the evening can increase the efficiency of a solar panel up to 45%, according to some manufacturers [Linak]. Precise tracking of the sun is achieved through systems with single or dual axis tracking.

Watch this introductory video (5:33), which provides an illustration to the benefits of sun tracking:

DEGERenergie - Solar Tracking Systems (5:33)

Systems that employ trackers

So, what types of systems should include tracking devices (a.k.a. trackers)?

First of all, the systems that specifically utilize the direct beam radiation benefit from tracking. In majority of concentrating solar power (CSP) systems, the optics accept only the beam radiation and therefore must be oriented appropriately to collect energy. Such systems will not produce power unless pointed at the sun. Tracking is required for heliostats in central receiver (solar tower) systems. CSP collectors require significant degree of accuracy of sun tracking.

In photovoltaic (PV) applications, tracking devices can be used to minimize the angle of incidence of incoming solar rays onto a PV panel. This increases the amount of energy produced per unit of installed power generating capacity. This increases the efficiency of the system and its cost-effectiveness, but, at the same time, tracking is not strictly required for regular flat panel PV as they accept both beam and diffuse radiation.

In concentrating photovoltaics (CPV), the optics requires beam radiation and therefore must be oriented appropriately to focus light on the PV collector to maximize the energy converted. CPV modules that concentrate in one dimension must be tracked normal to the sun in one axis. CPV modules that concentrate in two dimensions must be tracked normal to the sun in two axes [Solar Tracker from Wikipedia.org]. CPV modules require high degree of accuracy of sun tracking.

Single-axis and Dual-axis

There are many types of solar trackers, which are different in costs, design complexity, and performance. But we can distinguish two basic classes of systems:

1. Single axis trackers
   The single axis solar trackers can either have a horizontal or a vertical axis. The horizontal axis is used in tropical regions where the sun gets very high at noon, but the days are short. The vertical type is used in high latitudes, where the sun does not get very high, but summer days can be very long. In concentrated solar power applications, single axis trackers are used with parabolic and linear Fresnel mirror designs.
2. Dual axis trackers
   The dual axis solar trackers have both a horizontal and a vertical axis, and thus they can track the sun's apparent motion at any location. Dual axis tracking is commonly used for CSP applications, such as solar power towers and dish (Stirling engine) systems. Dual axis tracking is extremely important in solar tower applications due to the angle errors resulting from longer distances between the mirror and the central receiver located in the tower structure.

In more detail, these types of trackers will be studied in Section 3.3. of this lesson.

Pros and Cons

With tracking incorporated in the system design, the cost of the system is understandably higher compared to fixed tilt systems. According to the US DOE report [Barbose et al., 2013], "among projects completed in 2012, the capacity-weighted average installed price in US dollars was 3.3/W for systems with crystalline modules and fixed tilt, compared to 3.6/W for crystalline systems with tracking and 3.2/W for thin-film, fixed-tilt systems." Efforts are constantly made by manufacturers to lower the cost of the tracking systems, making them less complex, more compact, reliable, and easier to maintain. In spite of the additional costs, use of trackers is often a preferred option for utility-scale installations due to the significant boost to the system performance. Figure 3-1 shows the trend of increasing use of tracking systems in the U.S. utility-scale PV installations over the 2007–2017 decade. Cumulative tracking system installation reached 79% in 2017 (meaning that only 21% of large PV installations opt not to use trackers). These data include both one-axis and dual-axis tracking systems cumulatively, however there are many more one-axis trackers deployed than dual-axis trackers.

Figure 3.1: Percentage of U.S. utility-scale PV systems using tracking systems, 2007–2017

Source: NREL, U.S. Solar Photovoltaic System Cost Benchmark: Q1 2018

For most solar tracking applications, we need a reasonably accurate knowledge of where the sun will be at a specific hour during each day in a year. Theory is well-developed to calculate the sun position with respect to the observation point on the earth surface, and it sets the background for design and modeling of both photovoltaic and concentrating solar power systems of various scale.

In order to discuss tracking or any other adjustments of solar receivers, it would be useful first to understand the sun's path across the sky dome. We are going to turn to the following reading, which describes the key parameters of the solar motion.

The above materials provide the main tools for predicting the position of the sun at a location of choice at any specific time. Let us summarize a few key takeaways from this reading.

Solar Altitude and Solar Azimuth

The main parameters to determine are solar altitude (α) and solar azimuth (z). Here are the equations that are used to calculate these coordinates:

Let us consider an example showing how to use these equations.

This calculation can be essentially used for any location and any time in a year. The algorithms available help to produce detailed solar resource data for different settings. These data are available for reference and use, so you do not have to calculate all things from scratch, although it is useful to understand the theoretical background of it.

Sun Path Chart Tool

We can use the Sun Path Chart Program calculator at the University of Oregon's Solar Radiation Monitoring Laboratory website to obtain a complete picture of sun movement throughout the year. The calculator allows data to be plotted in either orthogonal or polar coordinates. For example, the diagram below (Figure 3.2) was obtained for the same location (Abu Dhabi).

Figure 3.2. The orthogonal projection of the solar path for the location.

Credit: (Abu Dhabi, 24.5o N, 54.4o W) obtained by the SunChart calculator [UO SRML, 2007].

In this diagram, the solar altitude (elevation) is plotted versus solar azimuth, as shown by the blue curves for each date. There are a few representative dates shown, and January 21 is the closest to the calculation example previously given. Note that the solar azimuth is given on the 360^o scale, with 180^o corresponding to the south. Alternatively, Kalogirou uses the coordinate system and formulae to calculate solar parameters versus 0^o as true south, with negative azimuth values corresponding to morning and positive azimuth values corresponding to afternoon hours. So beware of that difference if you try to match data from both sources. On the Sun Path diagram, the hourly position of the sun is marked by the red curves. In this particular case, the local standard times are plotted, while a similar diagram can be made in terms of solar time.

The above materials and activity make sure that you can employ proper tools for defining solar position on the sky dome. Further on, the receiver positioning algorithms will use this information as the operational basis. Different types of tracking systems are discussed in the next section of the lesson.

In technical sense, sun tracking is a method to keep the surface of the solar panel or a collector perpendicular to the incident solar rays. This is the ideal condition, when maximum amount of solar energy is transmitted to the receiving surface.

When the incident ray is not perpendicular to the surface (which is often the case with fixed-tilt systems), the angle of incidence is not zero (q ¹ 0), and part of the incident energy will be lost due to so-called cosine effect. To maximize efficiency of the system, we should always seek ways to minimize the cosine effect at any particular moment of time.

The figure below shows two scenarios: the left image illustrates an ideal situation, when solar rays come down on the surface of solar collector (PV panel) at the 90^o angle; the right image shows what happens when the Sun moves across the sky while the panel remains fixed.

Figure 3.3: Orientation of the receiver plane perpendicular to the incident rays (left) and at an angle (right), which introduces cosine effect.

Mark Fedkin

In the second case, the sun rays come down to the surface at an angle q, which will decrease the amount of energy absorbed by the surface, and thus will lower the system efficiency. By how much?

We can try to estimate this reduction due to cosine effect if we break down the G vector into two components: one perpendicular to the surface ($G_{\perp }$) – useful component that would be absorbed, and one - parallel to the surface ($G_{| | }$) – non-useful component that would be reflected or somehow lost.

Figure 3.4: Evaluation of cosine effect on a horizontal surface based on the zenith angle.

Mark Fedkin

For example, if we assume incident irradiance to be 1000 W/m² and angle of incidence 30°, then

$$G = 1000 \times \cos \left( {30}^{\circ }\right) = 866 \mathrm{W} / {\mathrm{m} }^{2 }$$

Thus, without considering other inefficiencies, losses due to cosine effect are expected to be around 13.4% at this angle, which is quite substantial.

Tracking can be an effective solution to minimize these performance losses. Tilting the panel by the angle (b) equal to the zenith angle would set the panel perpendicular to the sun rays once again.

Figure 3.5: Tilting of the receiver plane to avoid the cosine effect losses.

Mark Fedkin

The early attempts to eliminate the cosine effect would involve annual adjustment of panel angle throughout the day. But that would be tedious, inaccurate, and too discrete, while the Sun stays in constant motion on its daily path. Present-day automatic trackers use algorithms that are able to continuously track the Sun with an accuracy of $\pm {0.0003}^{\circ }$.

Tracking systems are classified by the mode of their motion. We can define three axes for a moving surface (which represents a receiver): two horizontal axes and one vertical axis (Figure 3.3). The surface can be rotated around each axis (tilted) to achieve an appropriate angle with respect to the incident solar beam. When movement or adjustment of the surface is done by rotating around one axis (tilting), it is single-axis tracking. When rotation of the surface is done around two axes simultaneously, it is two-axis tracking. Two-axis tracking allows for the most precise orientation of the solar device, is reported to provide 40% gain in energy absorption, but it is more complex and costly. Such two-axis systems are also used for controlling astronomical telescopes.

Figure 3.3. Three axes of rotation of a hypothetical moving surface.

Credit: Mark Fedkin

In case of single-axis tracking, the axis of rotation is usually oriented in the N-S direction or E-W direction. Tilting is performed in a way to minimize the incidence angle. In case of two-axis tracking, ideally, the incidence angle is always zero, i.e., the surface is kept perpendicular to the solar beam.

Read about various tracking modes in the following sources.

So, from reading these chapters, you now have quite a complete list of different ways of tracking and corresponding formulae to describe the relative position of the sun and inclined surface. The following activities will give you an opportunity to practice the basic calculations involved in two-axis and single-axis tracking. The first problem considers a simple case of two-axis tracking. As long as we know the solar coordinates, we can orient the receiver in that direction. But the tracking system that moves the plane needs precise input data, which we try to obtain here.

Problem 3-2: Dual-axis tracker data

The system is a heliostat with two-axis tracking: one vertical axis, and one horizontal (SN) axis. The goal is to determine the azimuth for the heliostat orientation and tilt angle for the horizontal axis at any time of the day to supply these data to the tracking system. A sketch of the collector is given below, and the blue line is the horizontal axis we want to tilt. The red line denotes the vertical axis, about which the collector can be rotated.

Figure 3.4: Heliostat diagram

Calculate and tabulate a set of Z_s-β data for every hour during the daylight period on March 21 at your chosen location. Feel free to use any available resources (solar path diagrams or appropriate equations) to determine the position of the sun.

In this calculation, we can assume that incidence angle on the surface of the collector will be zero at any moment.
Please provide references and explanation to your work.

This calculation will be submitted as part of Lesson 3 problem set.

The second problem on this topic considers the single-axis tracking case - one with horizontal NS axis and EW tracking (see Kalogirou's chapter, p. 69). In this case, the receiver has only one degree of freedom, so its motion is limited. We will not be able to reach the zero incidence angle, but we will try to minimize it in order to maximize the solar radiation on the plane.

Problem 3-3: Single-axis tracker data

Consider a flat collector with a fixed horizontal NS axis and tilting EW axis (see sketch below, side view). Because the NS axis is fixed, the surface azimuth (Z_s) is either -90^o when it tilted east, or +90^o when it is tilted west. The β angle defines the tilt, which is applied to minimize the incidence angle on the surface.

Figure 3.4: Flat Collector diagram

For your chosen location, determine and tabulate the surface position parameters (Z_s-β ) for every hour on March 21st. Feel free to use any available resource to determine the sun position. Make sure to provide references and explanation to your work.

This calculation will be submitted as part of Lesson 3 problem set.

Afloresm

Image credit: Afloresm via Flickr

The main elements of a tracking system include [Rockwell Automation, 2011]:

• Sun tracking algorithm: This algorithm calculates the solar azimuth and zenith angles of the sun. These angles are then used to position the solar panel or reflector to point toward the sun. Some algorithms are purely mathematical, based on astronomical references, while others utilize real-time light-intensity readings.
• Control unit: The control unit executes the sun tracking algorithm and coordinates the movement of the positioning system.
• Positioning system: The positioning system moves the panel or reflector to face the sun at the optimum angles. Some positioning systems are electrical, and some are hydraulic. Electrical systems utilize encoders and variable frequency drives or linear actuators to monitor the current position of the panel and move to desired positions.
• Drive mechanism/transmission: The drive mechanisms include linear actuators, linear drives, hydraulic cylinders, swivel drives, worm gears, planetary gears, and threaded spindles.
• Sensing devices: For trackers that use light intensity in the tracking algorithm, pyranometers are needed to read the light intensity. Ambient condition monitoring for pressure, temperature, and humidity may also be needed to optimize efficiency and power output.
• Limit switchesare used to control speed and prevent overtravel. The mechanical overtravel limits are used to prevent tracker damage.
• Elevation feedback is accomplished by either 1) a combination of limit switches and motor encoder counts or 2) an inclinometer (a sensor that provides the tilt angle).
• An anemometer is used to measure wind speed. If the wind conditions are too strong, the panels are usually driven to a safe horizontal position and remain in the safety position until the wind speed falls below the set point.

Three classes of tracker drive types to operate the moving receiver:

1. Passive trackers use the sun's heat to expand the compressed gas, which is used to move the panel. Selective heating of some cylinders versus others creates more expansion on one side of the panel and makes it tilt. These systems are relatively simple and low-cost, although they may lack the due precision necessary for the solar conversion systems using concentrated sunlight.
2. Active trackers use hydraulic or electric actuators to move the panel based on sensor response. Light sensors are positioned on the tracker at different locations for higher precision. These systems work best with direct sunlight and are less efficient with cloudy skies.
3. Open-loop trackers use pre-recorded data on the sun's position for a particular site. Simple timed trackers move the panel at discrete intervals to follow the sun's position but do not take into account the seasonal variations of the sun's altitude. The altitude/azimuth trackers employ astronomical data to determine the position of the sun for any given time and location.

Actuators

Linear actuators are common technical tools that proved to be effective solution for moving the solar receivers. An electric linear actuator is a device that converts the rotational motion of an electric motor into linear motion. With linear actuators you can lift, slide, adjust, tilt, push, or pull objects of various masses, and they are easy to implement in many different applications. Mechanically, linear actuators are quite simple devices that have been extensively deployed in 2-axis and 1-axis trackers due to their precision and service reliabilty.

The following video provides a rather detailed overview of the design, principle of operation, and specifications of electric linear actuators:

Video: Linear Actuators 101 (19:43)

The technical details of all the components of tracking systems would be beyond the scope of this course. It is important to understand, though, that additional components and more complexity, while improving the efficiency of the solar panels and reflectors, add to the cost of the whole system and consume additional energy.

This following video (4:25) demonstrates some technical features of a single-axis tracking system:

Video: Renewable Energy: Single-Axis Tracker (4:24)

In Lesson 3, we discussed the benefits of sun tracking for performance of the solar energy conversion systems. It is clear that although tracking helps to collect more solar radiation per square unit of solar receiver, the tracking systems may be complex and costly, and hence should be used only when benefits in terms of efficiency outweigh the expenses for extra energy and equipment. We reviewed the fundamentals of solar motion, and you should now be comfortable using the key equations to calculate the sun position at any time at any location on the earth. This lesson included description of different modes of tracking - single-axis and two axis - and gave you an opportunity to perform some basic calculations and work with available data on solar path. I hope you found the resources in this lesson useful and that, in the future, you will feel confident applying those calculation methods to the systems of your choice. Tracking certainly is a worthy technology when we look at the utility scale solar systems, as this technology provides an even more significant boost when scaled up. A number of companies are currently specialized in tracking technologies, constantly innovating and creating more and more robust systems for future solar plants.

The table below summarizes all activities that are due for this lesson. Some of those have been included in the body of the lesson, and this list simply repeats them for your reference.

References for Lesson 3

This lesson contains materials describing the main principles of photovoltaic conversion of light. You will learn what properties of PV materials determine the performance of the solar cells and how that performance is measured. You should be able to look at the performance curve and say if it is good or not if you deal with a single cell or a module. We will also see what happens to the parameters of the PV actions on system scale-up. Different types of PV materials and systems will be discussed here, and recent innovations and trends to improve solar cell efficiency will be reviewed.

Learning Outcomes

By the end of this lesson, you should be able to:

• understand and explain the principle of photovoltaic effect;
• analyze the photovoltaic cell performance data;
• discuss innovative ideas and recent trend in PV development.

The word "photovoltaic" immediately indicates the connection between light (phot- greek) and electricity (volt, unit for electric potential). The key property of a photovoltaic material is to convert light energy to electric current. This conversion takes place due to the photovoltaic effect - a physical phenomenon in a semiconductor, which we are going to discuss next.

Semiconductors are a special class of materials, whose conductance is not permanent, but rather depends on the energy available to activate electrons in the crystal lattice. Crystalline silicon is a semiconductor material widely used in photovoltaics. It becomes conductive when the energy of the photons absorbed by the crystal surface is sufficient to raise the electron state from the valence band to the conduction band. This required amount of energy to excite an electron is defined as band gap. Band gap is an intrinsic property of semiconductors and eventually has a direct influence on the photovoltaic cell voltage. The following schematic (Figure 4.1) provides a demonstration of the band gap concept.

Figure 4.1. Schematic illustration of the band gaps in various materials. The vertical axis is the electron energy, and E_F is the position of the Fermi level.

Credit: Band Filling Diagram.svg from Wikimedia.org

In this picture, we can visualize the difference between different classes of materials: conductors (metals), semiconductors, and insulators. The valence and conduction bands in a metal overlap, so it does not take any significant energy to free the electrons. They are available for conduction as soon as the potential gradient is provided. In insulators, the gap between the valence and conduction bands is very large, so it requires so much energy to free the electrons that it can damage the material itself. For semiconductors, the situation is somewhere in between. The band gap is big enough to prevent spontaneous conduction and to provide separation of charges, and small enough to be matched by photon energy.

The band gap energies of several different materials are listed in Table 4.1 below. Some of those materials are more suitable for photovoltaic applications than others. How suitable the materials are for photovoltaic applications would be determined by how close the photon energy is to the band gap of the material, and if the energy of photon is sufficient to cover the band gap.

For example, if we have a photon with energy of 2 eV hitting silicon surface, 1.1 eV of that energy will be used to move an electron to the conduction band; the rest of the energy (0.9 eV) will be dissipated as heat. However, if a material with a greater band gap is used, for instance copper oxide, 2 eV is not enough to free the electron. We would need a higher energy photon there.

Based on these reflections, we see that low band gap materials (such as germanium), can be used to capture low energy photons (like those in the red and IR parts of the spectrum), and high band gap materials (e.g. copper nitride or gallium phosphide) can be used for using high energy photons. Combination of different materials in one system allows for more efficient use of available radiation. At the same time, if the band gap of the PV material is too small compared to the incident photon energy, a significant amount of energy will be converted to heat, which is not a good thing for PV cell itself. No matter how much higher the photon energy is compared to the band gap, only one electron can be freed by one photon. This is the reason for the limited efficiency of the photovoltaic cells.

The data in Figure 4.2 show how the maximum efficiency of a solar cell depends on the band gap. If the band gap is too high, most photons will not cause photovoltaic effect; if it is too low, most photons will have more energy than necessary to excite electrons across the band gap, and the rest of energy will be wasted. The semiconductors commonly used in commercial solar cells have band gaps near the peak of this curve, for example silicon (1.1eV) or CdTe (1.5eV). The Shockley–Queisser limit (33.7%) defined at the peak of the curve has been exceeded experimentally by combining materials with different band gap energies into tandem solar cells.

Figure 4.2. The prediction of the solar cell maximum efficiency as a function of material band gap by the Shockley–Queisser limit.

Credit: ShockleyQueisserFullCurve.svg from Wikipedia.org

Next, please refer to the following reading to learn more about the background of the photovoltaic effect:

Based on the above reading, we can summarize that the photovoltaic effect essentially includes three main steps:

1. Absorption of light (photons)
2. Generation of charge carriers
3. Separation of charge carriers between electrical contacts

When all these steps occur, the system is able to generate electric current (flow of charge carriers), which can do work.

Answer the following questions for self-check:

While photovoltaic effect readily takes place in a number of materials, the third step - separation of the charge carriers - is probably the trickiest from the technical point of view. For example, in a regular silicon crystal, when absorption of a photon induces the release of an electron from the valence to the conduction band, a hole (positively charged locus) is formed in its place in the crystal lattice. Further, the excited electron and the hole can recombine to release heat. This is not what we want! So doping is often used to modify silicon structure. For example, boron (B) can be included into the structure in place of silicon. Because boron has a valence of 3 (versus silicon's valence of 4), there is a "gap" in the structure, which can accept an electron. This type of semiconductor, which has positive centers to accept electrons, is p-type semiconductor. Alternatively, silicon can be doped with phosphorus (P), which has a valence of 5 and brings an extra valence electron, which is not involved in covalent bonds. This electron can be donated. This type of semiconductor is n-type semiconductor.

Thus, the p-semiconductors have an excess of positive charge carriers, and the n-semiconductors have excess of negative charge carriers. If p- and n-types are put together, the interface between them will represent the p-n junction.

To understand how the p-n junction works, please watch the following video (10:36):

Video: The PN Junction. How Diodes Work? (10:36)

As we understand from the video, the p-n junction creates an internal electric field due to diffusion of charge carriers between two types of semiconductors (Figure 4.3). In this diagram, p-semiconductor is on the left, and n-semiconductor is on the right. p-semiconductor has excess of holes (positive charge carriers), as seen from the high position of the red curve on the left side, and the n-semiconductor has excess of electrons (negative charge carriers), as seen by the high position of the blue curve on the right side. At the junction region, which is defined as space charge region, there is a zone depleted of charge carriers: negatively charged impurities (shown as blue circles) "push away" the electrons in the n-semiconductor, and similarly, the positively charged impurities (shown as red circles) "push away" the holes in the p-semiconductor. This creates charge carrier separation. Also, in the depleted central region, co-presence of the negative and positive impurity atoms on both sides of the p-n boundary creates an electric field, which maintains that separation.

Figure 4.3. A p-n junction in thermal equilibrium with zero bias voltage applied. Light red zone is positively charged. Light blue zone is negatively charged. Gray regions are charge neutral.

Credit: Mu301 via Pn-junction-equilibrium.png from Wikimedia Commons

When light shines on the surface of the p-n material, photons excite electrons into conduction band, thus creating an electron-hole pair. If this happens in the n-doped side of the p-n junction, the newly excited electron is driven away from the junction, and the hole is swept across the junction to the p-doped side. This separation of the electron-hole pair is achieved by the action of the electric field in the space charge region. Then providing the external circuit (a wire) between the p- and n- semiconductors, we can initiate movement of the electron from the n-doped side to the p-doped side, where it recombines with a hole. This photo-induced electric current is the usable energy that can be harvested.

Please see the animated explanation to the light-induced current on the PVEducation website.

In the previous section, we understood how the photo-induced electric current is be generated at the p-n junction due to photovoltaic effect. How can we estimate the magnitude of that electric current?

To answer this question, first let us define the electron traffic across the band gap as generation or light-induced current (I_L). Therefore, each photon absorbed is responsible for contributing one electron to the generation current inside the device. Hence, we can write:

where I_L is light-induced generation current, q is the electron charge, N is the number of photons absorbed, and A is the surface area of the semiconductor exposed to light. Logically, we see that the more photons are absorbed, the higher the generation current. Also, the greater the area of semiconductor exposed to light, the higher the generation current. To be independent of the size of the cell, we can express this relationship in terms of current density (J_L), which is current normalized by area:

For example, we can try to use this equation to estimate the current density of a photovoltaic device corresponding to the typical terrestrial light spectrum. Inputting values for electron charge (1.6 x 10^-19 C) and number of photons in the absorbed range of spectrum for crystalline silicon (4.4 × 10¹⁷) to the above equation, we obtain: $$I_L/A=qN=(1.6\times {10}^{-19})(4.4\times {10}^{17})=70{\text{ mA/cm}}^2$$

This is the maximum current density that could be expected from a silicon cell, if there were no losses, and if all of the electrons were perfectly transferred through the external circuit. In reality, current losses take place, so the actual measured current density will be less than that ideal value (Markvart, 2000).

Next, let us see how the solar cell voltage can be estimated. The maximum voltage of a solar cell is determined by the semiconductor band gap. The electrostatic energy available due to separation of electrons and holes cannot exceed the band gap energy; otherwise, recombination would occur. The cell voltage (V) upper limit is therefore set by the following expression:

Numerically, the maximum voltage in volts is equal to the band gap energy in electron-volts. For example, the maximum voltage for a silicon solar cell is $V_{max }= 1.1 V$

The same as with maximum current, the maximum voltage is never achieved practically because of losses and process limitations. However, in general, based on Eq. (4.4), the semiconductors with greater band gap indeed produce higher voltage (Markvart, 2000).

In the above description, for simplicity, we assumed that all the photons (with energy above the band gap) reaching the surface are absorbed and transfer their energy onto electrons. This would be ideal and would give us ideal generation current. As a matter of fact, absorption of photons by semiconductors is material dependent and is controlled by the absorption coefficient. This is an important property to take into account, since it directly affects the charge carrier generation rate.

At this point, we have already recognized that the key parameters describing the performance of a solar cell are current density and cell voltage. We have looked into their origin - how they develop in the cell due to the photovoltaic effect, and looked at some factors that affect that process. Now, we will proceed to examination of the I-V characteristic (a.k.a. performance curve) and see how it is obtained and what different parts of this curve tell us about.

Using electric circuit notation, a solar cell can be represented by a diode, which represents the p-n junction.

Figure 4.4. Equivalent circuit of a solar cell.

Credit: Mark Fedkin

The current through the diode (I_o) is the exchange current present when the element is in the dark. This is a small current compared to light-induced current (I_L), which passes through the external load. The net current is the difference between the light and dark currents, or including Shockley diode equation:

where V is cell voltage, q is the charge of electron, k is the Boltzmann constant, and T is absolute temperature. This is the main equation that describes the relationship between voltage and current in the solar cell in operation. If we plot the cell voltage versus current (or current density), we will obtain the curve that generally looks like this (Figure 4.5):

Figure 4.5. Example I-V characteristic of an operating photovoltaic cell module.

Credit: Mark Fedkin

There are several important conditions to note on this curve. We see that at current being zero, the cell has the highest voltage. Because there is no current, the cell does not produce any work, but the voltage magnitude indicates the potential of the cell to do work. This is the open circuit voltage (V_oc or OCV). The OCV is a very important characteristic of any galvanic cell (including solar cells), and it depends on the cell material. By re-arranging equation (4.5), and setting the net current to zero, we can express the open circuit voltage as follows:

At cell voltage set to zero, the cell current reaches some maximum limiting value, which is called short-circuit current (I_sc). This is the kinetic parameter that shows the maximum current the cell is able to generate. It depends on the number of photons being absorbed by the material, optical properties of the cell and its size. You can imagine that if the sunlight intensity decreases for any reason, we will see a decrease in the short circuit current for a particular device. In an ideal case, the short-circuit current is equal to light-induced current: I_sc = I_L.

At any point on this curve (in Figure 4.5), we can define power output as follows:

At some point, the power will reach its maximum point, and the current and voltage corresponding to that point are defined as maximum power voltage (V_mp) and current (I_mp):

$$P_{\max }=I_{mp}{V}_{mp}$$

These parameters are shown in the diagram in Figure 4.5. by blue dashed lines. They characterize the conditions when the cell produces the highest power output. This point is important because it is where the cell efficiency is usually determined.

At the maximum power point, we can also define the characteristic resistance of the cell (R_ch). If the resistance of the external load is equal to R_ch, then the maximum power is transferred to the load. The characteristic resistance can be determined from the Ohm's law:

The next term we need to define, when talking about the cell power output is the fill factor (FF). Please refer to the following reading to learn about the fill factor.

As you should have noted from the reading, the fill factor can be calculated as follows from the cell performance parameters:

The fill factor is a convenient metric to characterize the solar cell performance. For cells that work well, FF>0.7. Typical parameters of the single-crystal silicon solar cell are (Kalogirou, 2009):

$J_{s c }= 32 m A / c m^2$

$V_{o c }= 0.58 V$

$V_{m p }= 0.47 V$

$F F = 0.72$

$P_{max }= 2273 m$

Based on equation (4.10), the maximum power output of the PV system can be readily found using equation (4.11) if we know open-circuit voltage, short-circuit current, and fill factor.

$$P_{\max }=V_{oc}{I}_{sc}(FF)$$

The I-V characteristic is a convenient tool to explore the effect of various external variables on the cell performance. What is going to happen to a module output if temperature rises? What if light intensity drops because of the clouds? If cell is damaged or has bad contact with current collectors, how will it be reflected on the performance curve? Learn about these effects from the following readings:

As we understand from the previous sections of this lesson, the electricity output of a single solar cell is relatively small, so cells need to be combined to provide enough power for any applications. Cells can be connected in series or in parallel. For example, when two identical cells are connected in parallel, the voltage of the system remains the same as for a single cell, but the current is doubled. When the same two cells are connected in series, voltage is doubled, while current remains the same. Examples of these combinations are shown in Figure 4.6.

Figure 4.6. Performance I-V curves for two identical solar cells connected in parallel (left) and in series.

Credit: Mark Fedkin, adapted from Kalogirou, 2009

To provide significant power output, solar cells are typically grouped into modules. A module is an engineered system consisting of multiple solar cells, wiring, frame, and glass. A module is a typical commercial stand-alone unit for solar cell applications. Each module can consist of a variable number of cells arranged in two-dimensional structure (Figure 4.7). In one direction, cells are connected in series to a branch. Then several branches are connected in parallel to complete a module (Figure 4.7).

Figure 4.7. Connection of single solar cells into a module.

Credit: adapted from Kalogirou, 2009

In Figure 4.7, the module voltage is denoted V^M and module current is denoted I^M (respectively, the single cell voltage and current will be denoted V^C and I^C in this presentation). This scheme depicts an example module that consists of N_SM single solar cells connected in series and N_PM parallel branches (subscript SM stands for series connection within module, and PM - for parallel connection within a module). The modules differ in design and may have variable number of parallel branches in order to deliver a certain current level. Many commercial modules may have simply one branch, i.e., all cells connected in series. To describe the electrical performance of a solar module, we can look at the model described in the reading material referred below. The model allows estimating the generated current and voltage of a module based on I-V characteristics of single cells.

The next step up the size scale will be grouping modules into an array (the same as it was done with grouping single cells into a module). A number of modules are connected in series within each branch, and then several branches are connected in parallel to form an array. Arrangement of modules in an array is shown in Figure 4.8.

Figure 4.8. Connection of solar modules in an array.

Credit: adapted from Kalogirou, 2009

In Figure 4.8, each element is essentially a thumbnail of the arrangement shown in Figure 4.7. The total array current is I^A depends on how many branches the arrays has, and the array voltage V^A depends on how many modules are connected in one branch. Modules are common at the scale of distributed power generation (e.g. rooftops) and are also used as unit elements of larger solar facilities. Arrays are more typical for utility scale solar farms and plants. Next, you will learn how to calculate the voltage and current for large arrangements and thus to estimate the order of magnitude of power output.

Proceed to the following reading to become familiar with the basic models that describe the performance of solar modules and arrays.

Answer the following self-check questions to assess your learning of this section.

There are different types of photovoltaics, some developed long ago, and others that are relatively new. Descriptions below provide a brief overview of a few well-developed PV materials. As you read through, please also open the links within each paragraph to get more information about each technology.

Monocrystalline silicon

Monocrystalline silicon solar cells are probably the oldest type of solar cells. They are made from pure silicon crystal, which has continuous lattice and almost no defects. Its properties provide for high efficiency of light conversion (typical ~15%; recent developments by SunPower boast improved efficiencies up to 22-24% ). Manufacturing of the Si crystals is rather complicated, which is responsible for high cost of this type of photovoltaics. Recent developments have decreased the total thickness of Si material used in monocrystalline cells to reduce cost. The monocrystalline silicon cells have a typical black or iridescent blue color. The monocrystalline silicon cells are believed to be very durable and last over 25 years. However, their efficiency will gradually decrease (about 0.5% per year), so replacement of operating modules might be needed sooner. The main disadvantages of the monocrystalline silicon panels are high initial cost and mechanical vulnerability (brittle). (Solar Facts and Advice: Monocrystalline Silicon, 2013)

Polycrystalline (or multicrystalline) silicon

Polycrystalline cells are made by assembling multiple grains and plates of silicon crystals into thin wafers. Smaller pieces of silicon are easier and cheaper to produce, so the manufacturing cost of this type of PV is less than that of monocrystalline silicon cells. The polycrystalline cells are slightly less efficient (~12%). These cells can be recognized by their mosaic-like appearance. Polycrystalline cells are also very durable and may have a service life of more than 25 years. The cons of this type of PV technology are mechanical brittleness and not very high efficiency of conversion. (Solar Facts and Advice: Polycrystalline, 2013)

Amorphous silicon (Thin-film)

Thin film photovoltaic cells are produced by depositing silicon film onto substrate glass. In this process, less silicon is used for manufacturing compared to mono- or polycrystalline cells, but this economy comes at the expense of conversion efficiency. Thin-film PV have efficiency of ~6% versus ~15% for single crystal Si cells. One way to improve the cell efficiency is to create a layered structure of several cells. The main advantage of the thin-film PV technology is that the amorphous silicon can be deposited on a variety of substrates, which can be made flexible and come in different shapes and therefore can be used in many applications. The amorphous silicon is also less prone to overheating, which usually decreases the solar cell performance. Amorphous silicon is most developed among the thin-film PV. (Solar Facts and Advice: Thin Film, 2013)

Figure 4.9, below, shows the trend of development and commercial implementation of different types of silicon PV technologies. There is no clear domination of a specific type of silicon substrate, as they all present a trade-off between cost and efficiency.

Figure 4.9. Market share of different silicon-based PV technologies from 1990 to 2013.

Credit: Global Market Share by PV Technology from 1990 to 2013 from Wikimedia Commons

Cadmium Telluride, CdTe (thin-film)

CdTe PV are another kind of thin-film solar technology. It has become quite popular due to the lower cost per kW-hour. The best efficiency obtained with CdTe cells is around 16%. One of the advantages of the CdTe cells is that they capture shorter wavelengths of light than silicon cells can do. There are some environmental concerns related to the limited supply of tellurium and potential toxic impact of cadmium at the stage of CdTe panel disposal. Developing effective closed-loop recycling technologies can be a game-changing factor in favor of this technology. (Solar Facts and Advice: Cadmium Telluride, 2013)

Copper Indium Gallium Selenide (CIGS)

CIGS PV have become a popular new material for solar cells, as it does not contain toxic Cd, and has higher efficiency (just under 20%). At this moment, the CIGS are the most efficient among the thin-film PV technologies. While lab results confirmed high promise of this kind of photovoltaics, the mass production of CIGS proved to be a problem. The CIGS cells are manufactured by thin film deposition on a substrate, which can also be flexible (unlike the silicon cells). Similar to CdTe cells, the CIGS cells demonstrate good resistance to heating.

Polymer and organic PV

Organic materials are quite attractive since they can be involved in high-output manufacturing and also because they can be made in various thicknesses and shapes. These types of cells are relatively lightweight (compared to silicon cells). Also, they offer flexibility and relatively low fabrication costs. They, however, are much less efficient (about 1/3 of a typical Si cell efficiency) and sometimes prone to quicker degradation (shorter service life). More technical details about this type of PV technology can be read in this Wikipedia article: Organic Solar Cell.

These are a few most well-known varieties of PV technology, but there are many more innovations that are at the research and development stage. Breakthroughs in new materials and cell design may be responsible for the growth of the PV industry in the upcoming decades. For this lesson's discussion forum, you will be asked to do a search of recent innovations and share a quick synopsis of one that you believe is especially interesting.

This lesson intended to provide you with the key concepts and terminology on the photovoltaic light conversion. It covers the basics of the photovoltaic effect and physical phenomena behind it, although it does not go into the depth of the science, but rather focuses on the practical understanding. Materials in this lesson also provide explanation to how the performance of the solar systems - cells, modules, arrays - is measured. Now, you should have an idea what the main parameters of the solar cell performance are, how to obtain them, and interpret them. You have learned how some of the key environmental parameters, such as temperature, light intensity, etc., affect the PV cell performance. The question of combining PV cells into bigger systems is discussed and illustrated with some examples, and we will return to the analysis of the larger-scale solar systems in Lesson 6, when more details will be given.

References for Lesson 4

Concentrating photovoltaic systems use lenses or mirrors to concentrate sunlight onto high-efficiency solar cells. Light concentration increases the flux of photons to the surface, which increases the photovoltaic current dramatically and opens ways to raise the conversion efficiency. There are predictions that concentrating photovoltaics (CPV) will be the next big trend in solar technology, although the price of electricity delivered by CPV systems is still too high to be commercially competitive. In this lesson, we will study the principles of concentrating photovoltaic systems and see how the concentration affects different parameters of solar cells. Also, we will review materials used for manufacturing concentrating photovoltaics. Finally, in this lesson, we will turn to the examples of recently commissioned CPV plants, some of them reaching the scale of multi-megawatt power generating facilities.

Learning Objectives

By the end of this lesson, you should be able to:

1. understand and explain the principle of concentrating photovoltaics (CPV);
2. analyze the effect of light concentration on PV cell performance;
3. discuss examples of CPV applications on utility scale.

One of the ways to increase the output from the photovoltaic systems is to supply concentrated light onto the PV cells. This can be done by using optical light collectors, such as lenses or mirrors. The PV systems that use concentrated light are called concentrating photovoltaics (CPV). The CPV collect light from a larger area and concentrate it to a smaller area solar cell. This is illustrated in Figure 5.1.

Figure 5.1. This is one of the common types of concentrator cells based on Fresnel lens, which takes the parallel beam of sunlight and directs it to a small area. For an effective use of the Fresnel CPV system, two-axis sun tracking is needed to ensure that the rays are perpendicular to the lens.

Credit: Mark Fedkin

Lower efficiency CPV technologies may employ silicon, CdTe, and CIGS (copper indium gallium selenide) cells, but the highest efficiencies can be achieved with multi-junction cells. Field efficiencies for these multi-junction cells are in the 30% range, and laboratory tests have achieved upwards of 40% efficiency (Kurtz, 2011).

The CPV can only use direct beam radiation and cannot use diffuse radiation (diffused from clouds and atmosphere). Therefore, these systems are suited best for areas with high direct normal irradiance. For proper light concentration, sun tracking is needed for achieving high cell performance. Tracking is especially critical for high concentration systems. In general, the CPV can be classified into low-concentration, medium-concentration, and high-concentration.

The high concentration of sunlight achieved with multijunction cells requires more sophisticated cooling and tracking systems, which can potentially result in higher energy costs.

CPV technology is expected to grow and to expand on market. The cost effectiveness of CPV technology is related to the fact that much smaller sized solar cells are used to convert the concentrated light, which means that much less expensive PV semiconductor material is used. Also, the optics added to the system are made from glass and are usually less expensive than the cells themselves.

Figure 5.2. Photovoltaic power system on the roof of the St. Petersburg Academic University - Nanotechnology Centre of RAS. In the center, there is a typical design of the Fresnel CPV system represented by a module of multiple cells, with a separate Fresnel lens placed on top of each single cell. On the right side of the image, a regular PV silicon cell module is shown.

Credit: Brücke-Osteuropa - via Wikimedia Commons

CPV systems can produce significantly increased temperature on the surface of the PV material, so the energy should be distributed evenly over the cell area to avoid local overheating (hot spots), which can damage the material. Also, the thermodynamic efficiency of the photovoltaic conversion is less at elevated temperatures, so some kind of cooling may be beneficial. Active or passive cooling can be used. For the CPV cells with low and medium concentration ratios, active cooling is not necessary, since the temperatures reached are moderate. The high-concentration cells require high-capacity heat sinks to avoid thermal destruction of the materials.

Let us find out how the concentration of light affects the I-V characteristics of a solar cell. We remember from Lesson 4 that the generation current of a solar cell (I_L) is a function of number of photons (N) hitting the photovoltaic surface:

where q is the electron charge, and A is the surface area of the cell. When light is concentrated, the number of photons increases according to the optical concentration ratio, so does the cell current. So, for the short circuit current of a solar cell (I_sc), we can write:

where C_opt is the optical concentration ratio (its definition was covered in Lesson 3). For convenience, we can denote cell performance parameters at concentrated light with an asterisk:

This equation essentially shows how much the cell short circuit current will change when the available light is concentrated C_opt times. Then, we can substitute this equation to the I-V characteristic equation, which describes the cell performance over ranges of voltage and current:

where V_oc* is the open circuit voltage (at concentrated light), k is the Boltzmann constant, T is the absolute temperature, and I_o is the dark saturation current. Now, we are going to modify this equation because we want to find how the open circuit voltage at concentrated light would be related to the open circuit voltage at ambient light. We know that the short circuit current is the highest current a solar cell can show, while the dark current is a very low number, so the quotient in the parenthesis should be much greater than 1, and therefore, a simplified form of Equation (5.4) should be true:

Next, this equation can be modified by extending the natural log as follows:

The second term here is equal to V_oc - the open circuit voltage without concentration, so we can write finally:

From Equation (5.7), it is obvious that there is logarithmic dependence between the cell open circuit voltage and the light concentration ratio. For example, if $C_{o p t }= 10$, the $( \; k \; T \; \; / \; q \; ) \; \ln \; \; C_{ \; \; o \; \; \; p \; \; \; t \; \; }$ term would be equal to 60 milivolts at 25^oC - this is by how much the cell voltage will increase with tenfold light concentration. In case of higher concentration, for example, C_opt = 1000, the voltage increase would be expected to be closer to 178 mV at 25 °C, which is relatively modest compared to current increase.

To estimate the concentration effect on maximum power output, we will use the equation (which was introduced in Lesson 4):

Substituting here Equations (5.3) and (5.7) and re-arranging, we obtain:

$$P_{max }^{\ast }= P_{max }{C}_{o p t }( 1 + \frac{k T }{q}\frac{\ln C_{o p t }}{V_{o c }})$$

As you can see, the cell power can raise dramatically because of light concentration, mainly because the cell current is significantly increased.

From the maximum power equation, we can further derive the effect of concentration on cell efficiency:

In this equation efficiency, (η) is expressed as the ratio of maximum cell power output to the irradiance on the cell surface. So, for concentrated light, the irradiance will be amplified to G* (which is proportional to C_opt). The maximum power output at the concentrated light, P_max, can be expressed as V_oc*I_sc*FF according to equation (4.9) in Lesson 4. Therefore, the expression for efficiency at the concentrated light can be modified as follows:

The algebraic transformation above is done by substituting Equations (5.3) and (5.7) into the equation (you can check). As a result, we see how "concentrated" efficiency (η*) is related to "non-concentrated" efficiency (η) through the optical concentration ratio. Try to apply this equation to find out what happens with the efficiency if you concentrate light ten times:

As you can see, the efficiency of the solar cell increases slightly in concentrated light, but this increase is not as apparent as for absolute output parameters (e.g. power). This is because in efficiency we always consider a ratio of the output to input energy. Both output and input energies increase due to concentration, so based on Eq. (5.10) the efficiency does not change much. Moreover, the efficiency of real solar cells cannot increase indefinitely because of power losses to heat. The amount of those losses is determined by the cell series resistance (R_s). The higher the series resistance, the bigger the power losses:

Because the current flowing through the cell is proportional to the light concentration ratio, the power wasted can be presented as:

The power loss will grow very rapidly as the concentration ratio increases because of the exponent factor. So, there is no sense to increase concentration infinitely because those efforts may not pay off in terms of useful power increase. According to some studies (Luque, 1989), there is an optimum concentration ratio for each type of cells. It is pretty much dictated by the cell series resistance and can be expressed as follows:

Many solar cells designed for concentrated light in fact have special features to reduce the series resistance, but the limits of design may still be dependent on the cell material. For silicon, for example, it is hard to create cells that would be efficient at concentration ratios higher than 200 (Markvart, 2000).

Traditional PV systems use a large amount of silicon; in contrast, CPV systems use a small amounts of high‐efficiency PV materials. A typical example of such high-efficiency cells employed in high concentration CPV systems is a multijunction cell. The term multijunction refers to the cell structure, which has multiple p-n junctions combined within a single cell. Each junction is responsible for absorbing light within a particular wavelength range. All the junction currents are then combined to one output.

Combination of multiple p-n junctions within one cell is achieved by blending several semiconductor materials in layers or in other heterostructural formations. Manufacturing those formations can be tricky, and therefore costly. However, the pay-off on the efficiency side of the technology proves to be worthy. While single junction PV cells have the maximum theoretical efficiency around 34%, multiple junction cells can achieve in ideal case the limiting efficiency of 86.8% under concentrated sunlight (Wikipedia, Multijunction PV cell).

Consider the two single junction cases below (high band gap and low band gap) (Figure 5.3.1 and 5.3.2).

Figure 5.3.1. Low band gap cell produces high current (because a large portion of the spectrum is involved), but low voltage (because E_g is low). As a result, the access energy is lost to heat.

Credit: adapted from National Renewable Energy Laboratory (NREL) / S.Kurtz

Figure 5.3.2. High band gap cell produces high voltage (because E_g is high), but low current (because few photons will participate). The light below the band gap will be lost.

Credit: adapted from National Renewable Energy Laboratory (NREL) / S.Kurtz

Figures 5.3.1 & 5.3.2 show two scenarios that illustrate the use of low band gap (top) and high band gap (bottom) materials. They both have limitations, which are responsible for low efficiency of single-junction PV cells. This is why tuning the different bandgaps to different components of light is a good idea in terms of minimizing power losses. Specific disadvantage of each scenario is stated under the plots. Combining different materials with different band gaps allows the cell to absorb a wide range of wavelengths and thereby to reach the highest efficiency.

To combine different materials in the multijunction cell, certain requirements need to be met for lattice-matching, current matching, and providing high opto-electronic performance. Regarding the first requirement of lattice-matching, one needs to make sure that the lattice constants of materials included are close. If the mismatch is significant, intercrystal defects can lead to quick degradation of electronic properties. For current matching, ordering of semiconductor layers within the multijunction cell is done in such a way that high-bandgap materials are placed on the top of the "sandwich," while the low-bandgap materials are placed on the bottom (Figure 5.4). That allows the light with low energy (greater wave lengths) to transmit to the lower layers and to be usefully absorbed. This concept is demonstrated in Figure 5.4. In this design, the suitable bandgaps need to be chosen so that the currents generated at each p-n junction are matched. If, for example, one of the junctions produces much lower current, it will be detrimental to the total current of the cell (because the layers are connected in series).

Figure 5.4. The arrangement of different band gap materials within a multijunction PV cell.

Credit: NASA via Wikimedia Commons

Next, we will refer to the following review on multijunction cell materials, their materials and design.

Reading this article, note the basic principles of operation of multijunction cells and define the key parameters that are responsible for their high efficiency. The reading quiz in this lesson is largely based on this material.

CPV systems have been much less represented on market compared to traditional PV. In 2012, the only utility-scale CPV plant in operation was a 5 MW project in Hatch, New Mexico, (commissioned in June 2011) (Mendelsohn et al., 2012). However, the number of CPV projects launched for utility electricity production was rapidly growing. In 2012, CPV market was characterized by NREL as follows:

"The limited commercial success of CPV to date is partly due to the fact that these systems are more complex than PV systems. During 2008, as silicon prices were reaching new market highs, CPV systems appeared ready for a commercial breakthrough. Prices have since collapsed, however, and this has changed the economics of several alternative technologies, including CPV. Despite the dramatic decreases in silicon and conventional module pricing, the CPV market looks to be entering a tentative growth stage. According to NREL’s database, at least 10 utility-scale CPV projects, representing about 471 MW, are currently in development and hold long-term PPAs with utilities. San Diego Gas and Electric (SDG&E) holds the majority of these PPAs, both in terms of megawatts (410 MW, or 86% of total) and absolute numbers. One CPV project, the 30 MW Alamosa Solar Generating Project in Colorado, will be the largest CPV installation in the world when completed in 2012. Project developer Cogentrix received a DOE loan guarantee of $90.6 million in September 2011; this was the only loan guarantee awarded to a CPV project. Continued market growth for CPV will be the most important factor in keeping its costs competitive with traditional PV and with fossil fuels. Without manufacturing in the tens of megawatts per year, it is unlikely that CPV will achieve the cost reductions necessary to make it an economic technology, despite its high efficiencies" (Mendelsohn et al., 2012)..

Obviously, there is a certain degree of skepticism related to the economic viability of CPV utility-scale facilities, which currently rely on loans from government and investors. In spite of this fact, many energy analysts predicted fast growth of concentrated photovoltaics during the second decade of 21st century. The following web article talks about this trend based on some actual data. 

Recent Updates

The above analysis and predictions were made seven years ago. Since then, the PV market experienced rapid changes, and the rate and scope of those changes went beyond many predictions and proved to be disruptive to a number of current energy markets. CPV development has also been impacted. Let us take a look at a more recent NREL report that analyzed the status and promise of the CPV technology.

With the fast progress in research and development of concentrating photovoltaic technology, projects started to grow to implement CPV on the commercial scale. This section of the lesson introduces some examples of such implementations.

A summary of CPV projects now operating in the U.S. is given in Table 5.3 below. While CPV is less common in other world's locations, it would be worth to mention Golmud Plant in China (2012-2013), with two phases adding up to 137 MW capacity and Touwsrivier CPV Project in South Africa (2014), which is also one of the largest installations - 44.2 MW.

The Alamosa Solar Plant is one of the biggest project commissioned in the US, and represents one of the cover stories of CPV implementation. Some more details on this case are presented below.

Alamosa Solar Project

The Alamosa solar plant is located on 225 acres of land in Colorado and supplies electricity to the grid of the Public Service Company of Colorado. At the time of commissioning, Alamosa was the largest CPV plant in the world, but was later surpassed by the newly built plants in China. The plant boasts a set of the advanced controls to ensure grid efficiency. The loan issued on the project guarantees low risk profile, while it is clear that the CPV development still needs to provide lower electricity prices in the future to be long-term competitive with regular silicon PV and fossil fuel power plants. The news releases about the Alamosa plant are linked below.

Alamosa demonstrates the robustness and reliability of the Amonix CPV modules (Amonix 7700). The modules are grouped by seven into CPV systems (7 modules each). Every one of those systems has a separate inverter and controls.

Check out the design of the Amonix module in the following documents:

In this lesson, you learned about the special type of PV systems - concentrating photovoltaics. There are a few important features that make this technology attractive. They include:

• high efficiency of light conversion (reaching 40% and above based on some laboratory tests);
• small size of cells, which allows the use of less expensive PV materials;
• economic land use by CPV systems.

At the same time, sophisticated design, necessary for precise tracking and cell cooling, is responsible for higher cost of CPV systems and high price of electricity compared to regular PV or non-renewable power generation systems. The activities in this lesson are oriented towards understanding some fundamental parameters of PV and CPV system analysis.

The table below summarizes all activities that are due for this lesson. Some of those have been included in the body of the lesson, and this list simply repeats them for your reference.

This lesson considers the principles of connecting solar cells to modules, and modules to arrays and larger systems for utility-scale electricity generation. To ensure the high-efficiency operation of an array or plant, a number of system components and pieces of equipment are needed. Here you will learn about some of those items and their purpose. Some of the designer's questions are: How do we put those components together and properly connect them? How do we select and match the electrical parameters of those components? What electrical and civil works are required to build a large scale PV plant? We will see how these questions are answered for different scales and different structural variations of photovoltaic systems.

For grid-connected PV systems, such parameters as voltage, current, and frequency should be matched to the ranges used by the grid. DC voltage may need to be stepped up or stepped down to match the grid requirements, and finally, DC power needs to be converted to AC power. All these steps constitute power conditioning, which is performed by an inverter - a special device that is responsible for seamless integration of solar power into the electric grid. In this lesson, we will learn about different types of inverters, their operation principle, their role, and conversion efficiency. The activities for this lesson will include a discussion forum and a reading quiz. 

Learning Objectives

By the end of this lesson, you should be able to:

1. explain basic strategies for the design of large-scale PV plants;
2. name the key components of PV systems and explain their purpose and basic principle of operation;
3. list the main types of solar inverters and explain their role and efficiencies;
4. discuss the key issues related to the connection of PV systems to the grid.

The electric power generated by PV modules goes through a series of transformations before it reaches the grid. Those transformations specifically include adjustments of current and voltage, DC-AC conversion, and also distribution of power between storage and transmission paths. Cumulatively, we can call these operations power conditioning.

Power conditioning is an important function of any utility-scale solar plant, which ensures that the energy generated can be effectively and safely delivered to consumers. To accomplish the proper power conditioning, we need a number of specialized components (in addition to the PV modules), and we are going to take a closer look at some of those components and their operation principles in this lesson.

Photovoltaic plants contain a large amount of supporting equipment, which serves to balance the system and to make it sustainably operational. The extra components include inverters, controllers, transformers, wiring, connector boxes, switches, monitoring devices, charge regulators, energy storage devices - all of which help prepare electric power for utilization. PV systems are typically modular in design, so that additional sections can be added to the plant or removed for repairs without significant disruption of its infrastructure. The energy flow at the solar plant runs through a variety of devices, which are connected by wire network and related hardware. This supporting infrastructure is often referred to as balance of system (BOS). The quality of the BOS is very important for providing lasting and efficient operation. The industry goal is to provide PV systems with an operational lifetime of at least 25 years. [Foster et al., 2010].

The general line-up of the key components of the BOS is illustrated in the generic system scheme below.

Figure 6.1. Typical energy transformation path in a grid connected PV system.

Credit: Mark Fedkin

Let us briefly discuss the main components in this scheme and describe their functions.

Charge controllers or regulators manage the flow of electricity between the solar modules (arrays), energy storage, and loads. The appropriate charge control algorithm and charging currents need to be matched for the batteries (or other energy storage devices) used in the system. The main purposes of a charge controller is to protect batteries from damage and to prevent overcharging or excessive discharging. Typically, these devices operate in the switch on / switch off mode. For example, when the terminal voltages supplied from a PV system to the battery increases above a certain threshold value (V_max^off), the switch disconnects the PV array. The array is connected again when the terminal voltage drops below a certain value (V_max^on). This hysteresis cycle protects the battery from overcharging. Similarly, charge controllers help prevent battery excessive discharging. When the current of the load connected to the battery is higher than the current delivered by the PV array, the load is disconnected as the terminal voltage falls below V_min^off and is connected again when the terminal voltage increases above a certain threshold V_min^on. Charge controllers also participate in voltage conversion and maximum power tracking [Kalogirou, 2009].

Inverters - devices that convert DC power coming from the solar modules to AC power (necessary for grid) are critical components of any PV systems. Inverters convert DC power from the batteries or solar modules into 60 or 50 Hz AC power. As with all power system components, the use of inverters results in energy losses due to interferences. Typical efficiency of an inverter well matched to the array is around 90%. Inverters are key components in both grid-connected and distributed power applications, and usually are a significant part of system cost. The AC current produced by inverters can have square, modified sine, and pure sine wave output (Figure 6.2). The pure sine is high cost and has the best power quality. The modified sine is medium cost, but has less efficiency. The square wave is low cost and lowest efficiency, which is only used by some applications. Square wave signals can be harmful to certain electronics due to high-voltage harmonic distortion. Inverters are common sources of electromagnetic noise, which can interfere with sound and video equipment. So, the inverters boxes must be grounded according to the code requirements and safety reasons [Foster et al., 2010].

Figure 6.2. Inverter output types.

Credit: Mark Fedkin after Foster et al., 2010

Grid-tied inverters are used to tie the PV system to the utility grid. They convert DC power to AC power in synchronization with the grid. For example, if grid fails for any reason, the inverter will shut down as well. The main considerations related to PV-grid interconnection include safety, power quality, and anti-islanding. Islanding is the condition when in case of power grid going down, inverter attempts to power the grid and can result in equipment damage and safety risks to technical personnel. Grid power line with PV modules connected to it is a typical islanding situation. After the grid is down, the PV panels still continue to power the line as long as the solar radiation is present. Thus, we have an "island" of powered line within un-powered grid. Most AC grid-ties inverters have anti-islanding feature, so the inverter will reduce power to zero within 2 seconds of the grid shut-down.

Inverters are rated by the total power capacity (from hundreds to a million watts). Some inverters have a good surge capacity for starting motors, but others may have limited surge capacity. So, designers should specify both type and size of the load to be connected to the inverter [Kalogirou, 2009].

DC-DC converters or transformers are used to step up (boost) or step down (buck) voltage of DC current. Therefore, the voltage of the solar array can be chosen independently of the voltage of the load. This kind of convenience comes with a cost - DC-DC converter always have losses, although the good models have efficiency as high as 95%, with some waste heat generated [Mertens and Hanser, 2013]. Ideally (if there were no losses):

$$P_{1 }= V_{1 }\times I_{1 }= V_{2 }\times I_{2 }= P_{2 }$$

Where V₁ and I₁ are voltage and current at the input (from solar module) and V₂ and I₂ are voltage and current at the output, respectively.

Batteries are used in many types of PV systems to supply power at low sun conditions (night or low irradiance). Additionally, batteries are required in solar systems because of the fluctuating nature of the PV output. The battery size/capacity is selected according to the load. They are usually connected in parallel to match higher capacity. There are several types of batteries commercially available for solar applications, including lead-acid, nickel-cadmium, nickel hydride, and lithium-ion. The main requirement for the batteries that are used as energy storage for solar systems is that they must be able to go through deep charging and discharging cycles without too much degradation. Batteries are classified by the nominal capacity (q_max), which is the maximum number of ampere-hours that can be extracted from the battery under certain standard conditions.

The efficiency of a battery can be defined as the ratio of the charge extracted during discharge to the amount of charge needed to restore that state of charge. The efficiency depends on State of Charge (SOC), which is the ratio between the present capacity of the battery and the nominal capacity (SOC = q/q_max). For example, SOC=1 when the battery is fully charged, and SOC=0 when the battery is fully discharged. The battery lifetime is often presented as the number of charge-discharge cycles the battery can sustain before losing 20% of its nominal capacity [Kalogirou, 2009].

Batteries used in power generating solar systems are actually different from car batteries. The car batteries are not designed to withstand the deep charge-discharge cycles and therefore should not be used with solar power generation. Batteries are usually installed in well-ventilated locations - (e.g., utility rooms) to minimize hazards from spills and made them available for easy maintenance or replacement. More details of energy storage technology will be covered in Lesson 9.

Grounding and bonding of related DC and AC circuits is important to maintain system integrity. According to the U.S. electric code, the systems operating under 50 V are not required to be grounded, although chassis grounding is required for all hardware.

Power output from the solar arrays would be dependent on the arrangement and connection of the modules within the plant and can be varied based on local preference. To know how we can manipulate power output, we need to understand the principles of interconnections of PV cells in the modules and connections of those modules to the direct current equipment. From the following reading, you will learn about different types of connections and how they determine the module response to disturbances such as shading or formation of hot spots. Additionally, we will see how the module performance may be affected by different environmental parameters, such as irradiance, temperature, and type of solar cell material. Finally, this reading will describe how the cable connections are made between the solar modules and a generator.

Note: The quiz at the end of this lesson will include a few questions on this reading. Please refer to the Summary and Activity page for more details.

Development of large solar PV plants has been underway in a number of countries, especially where the government-backed incentives and legislation were in place to support renewables. Since 2000, installation of MW-scale PV systems has been initiated in Germany, Spain, Italy, Greece, and further taken into even larger scale in U.S., China, India, and Brazil. The trend really picked up after 2007, when the number of the more than 1 MW systems grew from 20 to over 100 within a few years. Industry experts predict that the trend will hold and perhaps even accelerate into the future as the demand for renewable energy resources escalates. The average and maximum size of the utility solar plants (typically 5 MW to 500 MW) increases as well. The Topaz Solar Farm, located in San Luis Obispo County, California, is one of the largest solar photovoltaic power plant in the U.S. (Figure 6.3). This facility has the capacity to generate 550 megawatts (MW) of solar electricity using 5 million panels.

Figure 6.3. Space image of the Topaz Solar Farm (left) [image source: NASA, public domain] and installation view (right).

Credit: Pacific Southwest Region via Wikimedia

The list of the world's largest PV plants is updated from year to year, and you can see that Topaz benchmark has been already beaten repeatedly now, and the top facilities currently exceed GW limit. One of the attractive factors of the utility PV plants is that those facilities can be built relatively quickly (within 6-12 months) due to modular structure, unlike major hydroelectric, geothermal, or fossil fuel facilities, which would be typically developed over 3-5 year span. This presents a great opportunity for emerging economies to effectively meet their growing energy demands, especially since many of those countries possess an excellent solar resource.

Initially, the PV plant design is developed at the stage of feasibility assessment, which includes estimation of solar resource and expected yield. Then, the plant design is further improved, taking into consideration other local limitations and constraints. The feasibility stage also includes site measurements, topography mapping, environmental setting assessment, and social impacts. Key design features include such technical information as PV module type, tilting angle, mounting and tracking systems, module arrangement, and balance of system (BOS) components - inverters, connections, switches, and storage solutions. Further optimization of plant design would deal with such issues as shading, performance degradation, and economic trade-offs between increased complexity and energy yield.

The design of a utility scale PV plant is a complex endeavor. With many available choices of components and options for optimizing performance, it is important to strike a balance between cost savings and quality. Engineering decisions require significant technical expertise and should be "informed" decisions based on both optimization models and practical experience.

The following reading will introduce the main principles of the design of very large PV systems.

Following this reading, please take the reading quiz in Canvas (see Module 6).

Now, let us zoom in and take a closer look at the one of the key components of power conditioning chain - inverter. Almost any solar systems of any scale include an inverter of some type to allow the power to be used on site for AC-powered appliances or on the grid. Different types of inverters are shown in Figure 11.1 as examples. The available inverter models are now very efficient (over 95% power conversion efficiency), reliable, and economical. On the utility scale, the main challenges are related to system configuration in order to achieve safe operation and to reduce conversion losses to a minimum.

Figure 11.1. Inverters: small-scale inverter box for residential use (left) and Satcon utility-scale inverters (right)

Credit: Lauren Wellicome and Darin Dingler via Flickr

The three most common types of inverters made for powering AC loads include: (1) pure sine wave inverter (for general applications), (2) modified square wave inverter (for resistive, capacitive, and inductive loads), and (3) square wave inverter (for some resistive loads) (MPP Solar, 2015). Those wave types were briefly introduced in Lesson 6 (Figure 11.2). Here, we will take a closer look at the physical principles used by inverters to produce those signals.

Figure 11.2. Different types of AC signal produced by inverters.

Credit: Mark Fedkin

The process of conversion of the DC current into AC current is based on the phenomenon of electromagnetic induction. Electromagnetic induction is the generation of electric potential difference in a conductor when it is exposed to a varying magnetic field. For example, if you place a coil (spool of wire) near a rotating magnet, electric current will be induced in the coil (Figure 11.3).

Figure 11.3. Schematic illustration of electromagnetic induction

Credit: Mark Fedkin

Next, if we consider a system with two coils (Figure 11.4) and pass DC current through one of them (primary coil), that coil with DC current can act analogously to the magnet (since electric current produces a magnetic field). If the direction of the current is reversed frequently (e.g., via a switch device), the alternating magnetic field will induce AC current in the secondary coil.

Figure 11.4. Inverter cycles. During the 1st half cycle (top), DC current from a DC source - solar module or battery - is switched on through the top part of the primary coil. During the 2nd half cycle (bottom), the DC current is switched on through the bottom part of the coil.

Credit: Mark Fedkin

The simple two-cycle scheme shown in Figure 11.4 produces a square wave AC signal. This is the simplest case, and if the inverter performs only this step, it is a square-wave inverter. This type of output is not very efficient and can be even detrimental to some loads. So, the square wave can be modified further using more sophisticated inverters to produce a modified square wave or sine wave (Dunlop, 2010).

To produce a modified square wave output, such as the one shown in the center of Figure 11.2, low frequency waveform control can be used in the inverter. This feature allows adjusting the duration of the alternating square pulses. Also, transformers are used here to vary the output voltage. Combination of pulses of different length and voltage results in a multi-stepped modified square wave, which closely matches the sine wave shape. The low frequency inverters typically operate at ~60 Hz frequency.

To produce a sine wave output, high-frequency inverters are used. These inverters use the pulse-width modification method: switching currents at high frequency, and for variable periods of time. For example, very narrow (short) pulses simulate a low voltage situation, and wide (long pulses) simulate high voltage. Also, this method allows spacing the pulses to be varied: spacing narrow pulses farther apart models low voltage (Figure 11.5).

Figure 11.5. Pulse-width modulation to approximate the true sine wave by high frequency inverter.

Credit: Mark Fedkin modified after Dunlop, 2010

In the image above, the blue line shows the square wave varied by the length of the pulse and timing between pulses; the red curve shows how those alternating signals are modeled by a sine wave. Using very high frequency helps create very gradual changes in pulse width and thus models a true sine signal. The pulse-width modulation method and novel digital controllers have resulted in very efficient inverters (Dunlop, 2010).

The efficiency of an inverter indicates how much DC power is converted to AC power. Some of the power can be lost as heat, and also some stand-by power is consumed for keeping the inverter in powered mode. The general efficiency formula is:

where P_AC is AC power output in watts and P_DC is DC power input in watts.

High quality sine wave inverters are rated at 90-95% efficiency. Lower quality modified sine wave inverters are less efficient - 75-85%. High frequency inverters are usually more efficient than low-frequency.

Inverter efficiency depends on inverter load.

Figure 11.8. Typical generic inverter efficiency curve. Below 10-15% of power output, efficiency is quite low. At high output power, the efficiency is steadily high with some small variations.

Credit: Mark Fedkin

The behavior in Figure 11.8 partially results from the fact that stand-by losses for an inverter are the same for all output power levels, so the efficiency at lower outputs is affected more.

There are three types of efficiency ranking used for inverters. You may come across those numbers as you research different models and manufacturers. Those three types are:

1. Peak efficiency (shown by arrow in Figure 11.8) indicates the performance of the inverter at the optimal power output. It shows the maximum point for a particular inverter and can be used as a criterion of its quality.
2. European efficiency is the weighted number taking into account how often the inverter will operate at different power outputs. It is sometimes more useful than peak efficiency, as it shows how the inverter performs at different output levels during a solar day.
3. California Energy Commission (CEC) efficiency is also a weighed efficiency, similar to the European efficiency, but it uses different assumptions on weighing factors.

The main difference between the European and CEC efficiencies is that the assumptions about the importance of each power levels for a particular inverter are based on the data for Central Europe in the former case, and California in the latter. Hence, different formulae are used to calculate those values:

These methods of calculations need to be taken into account when using inverter specifications (Martin, 2011).

To learn more details about inverter efficiency, please go to the following reading.

Please answer the following self-check questions based on the above material.

Switching function in inverters is needed to alternate the direction of the DC current in order to produce AC power. Usually, electronic semiconductor devices are used to perform switching, such as transistors and thyristors.

Thyristors are used in basic models of inverters. They have three leads and usually "switch on" in response to current applied to one of the leads. Thyristor have only two modes: ON and OFF, the same as mechanical switches. More details on thyristors can be found on this Thyristor Wikipedia Page.

Transistors are similar in switching capability to thyristors, but they instead respond to voltage applied rather than current. That allows to smoothly vary the transistor's internal resistance. So in addition to ON and OFF functions, transistors also allow dimmer capability. More details on transistors can be found on this Transistor Wikipedia Page.

There are two main types of transistors used in solar inverters:

• Metal Oxide Semiconductor Field Effect Transistors (MOSFETs)
• Insulated Gate Bipolar Transistors (IGBTs)

The MOSFET type is suitable for very high switching speeds (up to 800 kHz), but operate at relatively low voltage. The IGBT type switch at lower speeds (below 20 kHz), but withstand higher voltage and high current (Dunlop, 2010).

Switching Control

Switching devices, such as thyristors and transistors, need to be controlled by an external signal. In the basic inverter designs, switching is controlled by the utility power line. Such switching devices are referred to as line-commutated. They are turned on and off by alternating half-cycles of the utility voltage, thus synchronizing the inverter output with grid line current. Although efficient, the line-commutated inverters have one disadvantage: they cannot operate independently of the grid.

Some inverters may contain an internal device that controls switching. Such a device is usually a microprocessor that provides precise timing. Such inverters are called self-commutated. Self-commutated inverters have additional capabilities of shaping the AC output and suppressing harmonics. And they can operate independent of utility power. There are two varieties of self-commutated inverters: voltage-source and current-source. They take the DC input as voltage source or current source, respectively, for conversion of the power to the AC output. Most of the present day inverters involved in solar plants are self-commutated (Dunlop, 2010).

A typical output voltage of PV panels can be on the order of 30 V, and it is too low for being converted to AC and fed to the grid. Therefore, DC/DC conversion is often a necessary step before the DC current from the PV system is supplied to the inverter. Most of the power conditioning units include some type of DC/DC converter. Direct current converter transforms the DC voltage V₁ to DC voltage V₂ via adjusting the current (I):

This is an ideal case, when input power is equal to the output power. In reality, there are always conversion losses, which lead to typical efficiencies in the range 90-95%.

DC/DC conversion allows keeping the voltage on the PV and voltage on the load separately controlled. There are two main types of DC/DC converters depending on the direction of voltage change: (1) boost converters transform smaller voltage to higher voltage and (2) buck converters transform higher voltage to lower voltage.

Please proceed to the following reading to learn more details about the operating principles of the DC/DC converters.

MPP tracking

One of the important functions that DC/DC converting devices can perform is maximum power point (MPP) tracking. The idea behind it is to keep the solar power system operating constantly at the maximum power, i.e., at the voltage V_mp and current I _mp. These parameters were discussed in Lesson 4, and Figure 4.4 illustrates the concept.

Assume that at a certain ambient irradiance, a solar cell or an array operate at the maximum power. Then, if the irradiance conditions change, the performance characteristic (I-V curve) of the cell changes (Figure 11.6). Thus, if the output voltage is kept constant, the output current drops significantly. The MMP tracker is used to adjust the voltage to the new V_mp value, thus maintaining the maximum power output.

Figure 11.6. Solar Cell I-V Curve in Varying Sunlight
I-V performance curves of a solar cell in varying sunlight. The upper curves demonstrate the cell performance at higher irradiance conditions. The vertical path crossing the performance curves indicates the shift of the maximum power point.

Source: ZyMOS via Wikimedia Commons

In large solar facilities, it is beneficial to have an individual MPP tracking unit to be connected to each array output, since different arrays may have different I-V characteristics (due to varying irradiance, orientation etc.). This improves the overall performance of the plant.

An algorithm for MPP tracking is discussed in Section 7.1.3 of the above-referenced reading.

Please answer the following self-check questions before proceeding to the next section.

Interconnections in solar systems and their connection to the public grid are regulated by National Electrical Code®(NEC). The NEC is a nationally recognized standard for safe electrical installation and is routinely applied all over the U.S. It is intended for use by trained personnel and is applied to integration of all wiring, overcurrent protection, disconnects, grounding, and equipment regulations. Article 690 of NEC is specifically dedicated to solar photovoltaic systems, and article 490 is applied to large utility-scale systems (over 600 V). Importantly, the NEC addresses the circuit requirements for solar systems, such as maximum current and voltage.

Maximum Voltage Rating

The first condition for determining the maximum DC voltage is that it should be lower than the voltage limits defined for any component on the DC side of the system. The maximum DC voltage output (Vmax) from a PV system can be estimated using the following formula (Dunlop, 2010):

where V_oc is the open circuit voltage of a module at 25 ^oC, n_ser is the number of modules connected in series, and C_T is the temperature correction factor. The C_T factors account for the voltage increase with decreasing temperature and can be found in Table 690.7 in Article 690—Solar Photovoltaic (PV) Systems of NEC.

Maximum Current Rating

Per NEC code Article 690.8, which deals with circuit sizing and rating, there are two different PV circuits distinguished:

(1) PV source circuits - conductors between PV modules and to the common point of connection, i.e., junction box

and

(2) PV output circuits - conductors between the junction box and the inverter or DC loads

These two types of circuits are schematically shown in Figure 11.7. The PV source circuits and PV output circuits are rated differently with respect to the maximum current.

Figure 11.7. Different types of PV circuits distinguished within a solar plant. Blue lines show connections in the source circuit; red lines show connections in the output circuit.

Credit: Mark Fedkin

The maximum DC current rating for PV source circuits is considered at 125% of the sum of all short-circuit currents rating of all modules. This assumption is based on the fact that under enhanced irradiance conditions, modules can occasionally generate currents higher than nominal I_sc values. The maximum source current is determined for each single string.

The maximum DC current rating for PV output circuit needs to take into account all parallel strings, and in this case the source maximum current needs to be multiplied by the number of strings (n_par) involved in the system.

In summary, the cables within each string should be able to withstand currents of 6 A, but the cables on the inverter side should be ready for 18 A.

The estimates of the system maximum voltage and currents are key factors for choosing the inverters, determining wiring, conductor size, and required overcurrent protection.

Interconnection types

By the type of interconnections, there are several inverter types employed at different PV systems. They include stand-alone, interactive, and bimodal inverters.

The stand-alone mode does not involve grid connection, but rather uses the battery storage to collect the power from PV and convert it to AC for specific applications (Figure 11.8a). If the battery storage is depleted, the system becomes stressed.

The interactive mode does not use any energy storage, and the inverter serves as the interface between the PV and the utility grid (Figure 11.8b). In this case, power can flow in both directions. If the on-site power demand is higher than the amount supplied from the PV, the system can draw power from the grid. If the produced power is greater than the on-site power demand, the excess power is fed to the grid.

Finally, the bimodal connection (Figure 11.8c) combines both stand-alone and interactive options. Here, the energy storage provides backup for critical loads, while the excess power is fed to the grid, like in the interactive mode. If for any reason the grid loses power, this bimodal system uses a separate dedicated distribution panel to support the critical loads (such as computers, lighting, water pumping, etc.) (Dunlop, 2010).

Figure 11.8. Different types of inverters connected to PV system: (a) stand-alone mode, (b) interactive mode, (c) bimodal mode.

Click here for a text description of the figure above

A. PV to charge controller to battery bank to stand-alone inverter to distribution panel (to load 1,2,3)

B. PV to interactive inverter to distribution panel (to load 1,2,3 & critical) to electric meter to utility grid

C. PV to charge controller to battery band to bimodal inverter (to distribution panel to critical load) to distribution panel (to load 1,2,3 & critical) to electric meter to utility grid

Credit: Mark Fedkin

Please refer to the following reading to gain more insight in inverter operation in grid-connected systems.

Grid-connection challenges

The following are some known concerns arising from interconnection of different scale PV systems to the utility grid.

1. Islanding is the condition when the solar power facility keeps supplying power to the grid during grid outage. This is a serious safety hazard, since utility workers who repair the grid may be exposed to unexpected voltage present in the utility line. To prevent damage to personnel and equipment, all grid-bound inverters must be able to detect outages and block power transfer. Inverters with such capability are referred to as anti-islanding inverters. Bimodal inverters may remain in stand-alone mode of operation while being disconnected from the utility grid line.
2. Power quality is reflected in the several electrical performance parameters, such as voltage, frequency, harmonic distortion, noise, etc. Many loads and equipment connected to the grid are designed to operate at certain prescribed conditions and may not be able to withstand voltage fluctuations and other disturbances. Low-quality inverters can cause poor power quality, which can cause damage to the sensitive equipment, or create hotspots in transformers.
3. Phase disbalance can occur if single-phase inverters are connected to three-phase power systems. Solution to the mismatch may be connection of three small inverters, each to a different phase, or employing a single three-phase inverter.

Please answer the following self-check questions before proceeding to the next section.

This lesson addresses multiple issues related to the scale up and deployment of photovoltaic technology. PV arrays generate DC power, sometimes a lot of it. It takes a lot of hardware and engineering effort to use that power safely and efficiently. Here, we looked into the principles of connection of various components within PV plants, overviewed the key pieces of equipment necessary for plant operation, and became familiar with some engineering practices used in construction and servicing of very large scale PV systems. We also discussed some technical aspects related to transformation of the energy supplied by PV into the usable grid power. Although the technologies involved in power conversion and conditioning are not directly solar, but rather supporting systems, it is important to understand their types, role, and specifications.

Please complete the following activities to complete this lesson.

In this lesson, we overview various Concentrating Solar Power (CSP) technologies, with emphasis on those that have been proved effective supplying utility scale power. The most well-known ones are parabolic trough systems, parabolic dishes, and 3D central receiver systems, such as solar power tower. The main purpose of a CSP technology is to convert solar radiation into thermal energy, which is in turn used to drive a heat engine. CSP plants have been commercialized and operate in a number of countries as part of power infrastructure. According to Renewable Energy World, the current global power generated by CSP exceeds 1095 MW. It is definitely a growing industry which works towards cost-effective solutions with renewable power, and may possibly become a major player in electricity generation in the U.S. over the following decades.

Here, we also review different types of power conversion cycles that are used for conversion of solar thermal energy to mechanical energy of the turbine. Such cycles are also called heat engines. They are crucial elements of any thermal plants, because in most cases mechanical energy or electrical energy are more practical than thermal energy. In this lesson, you will explore the principles and differences between several power conversion cycles and their varieties. One key question to answer will be which type of cycle would be suitable for a particular type of solar thermal plant and which parameters should be considered in that decision.

Learning Outcomes

By the end of this lesson, you should be able to:

1. Describe the principle of operation of the most important CSP technologies;
2. Apply performance metrics to characterize the utility scale CSP systems;
3. Understand the purpose and design of power conversion cycles;
4. Justify the choice of power conversion cycle based on solar system parameters and goals.

The term Concentrating Solar Power (CSP) covers a range of technologies that utilize optical devices, such as mirrors and lenses, to concentrate the beam solar radiation and to provide for higher efficiency of its conversion into other forms of energy. Typically, in many sources, CSP systems are associated with the solar thermal power; although, in a general sense, CSP can work with both thermal solar power and photovoltaic applications. Conversion of the concentrated sunlight can follow three routes: (1) conversion to heat, (2) conversion to electricity, and (3) conversion to fuels. The large utility scale plants primarily use the concentration of thermal energy, which is used to operate a steam turbine generator and produce electric power on site. In this case, the solar heat is used as any other source of heat (such as coal combustion, etc.) to generate steam.

The video below (11 min.) introduces the main types of the CSP technology. For that conversion cycle to run smoothly, day by day generating power, a few key technologies are linked together. These include light concentration, thermal transfer fluids, steam powered turbines, and sometimes thermal energy storage. The video also discusses the system efficiency at every step of conversion.  

Video: CSP Video Tutorial Unit 1-01 – Overview of CSP Technologies (11:21)

CSP systems can only benefit from the direct beam radiation and therefore are best suited to the regions with a high percentage of clear sky days. The locations that have significant cloudiness, smog, or dust are not favorable.

By concentrator configuration, the commercial CSP systems are represented by:

• parabolic trough systems
• central receiver towers
• parabolic dish systems
• linear Fresnel concentrators

These are the main technologies that you will read about in this lesson.

The CSP technology is one of the competitive options in energy industry for combustion-free electricity generation. Because the fuel cost is zero, the cost of the CSP technology is mainly associated with the significant initial capital investment. However, cost reduction trend for CSP is confidently predicted in the near future. The main avenues for cost efficiency are linked to:

• technical progress in CSP technology, research and development efforts, and lesson learned from operating plants; and
• scaling up to larger plant size, which allows for the use of more cost effective turbines for power conversion.

The CSP technology has been commercialized and has experienced rapid growth since 2005, in part stimulated by the international concern for increased fossil fuel combustion and climate change. Worldwide, the CSP expansion was led by Spain, which has most CSP plants currently installed. Over the past decade, the CSP sector experienced a significant slowdown due to the surge of photovoltaic systems in the renewable energy markets.

By technology type, CSP the current CSP market is led by the parabolic trough plants (over 75%). In spite of much slower growth than PV, solar thermal plants are nevertheless expected to be significant players in future energy economy. The main barrier to the CSP market growth remains high costs of electricity (average 0.20 USD/kWh compared to 0.05-0.10 USD/kWh for PV).

CSP market shares (based on data from fortunebusinessinsights.com) 

Mark Fedkin

Further, in this lesson, you will learn about different configurations of CSP plants, which differ in design of optical systems (reflectors and light collectors), the position of the receiver, and heat transfer networks. These various technologies have their pros and cons and are applied based on the target application, location, and other factors.

As was noted earlier in this course, parabolic trough technology is the most widespread among utility-scale solar thermal plants (Figure 7.1). The potential of this type of solar concentration is very high and can provide output fluid temperatures in the range 400-500°C. Parabolic trough is the linear-focus collector, which consists of a cylindrically curved parabolic mirror, which reflects the sunlight onto a tubular receiver positioned in the focus line of the parabola. The tubular receiver contains the fluid that absorbs heat and transfers it via circulation to the boiler or another device to produce steam.

Figure 7.1. Kuraymat parabolic trough solar plant, Egypt. The plant has the total solar aperture area of 130,800 m² and expected electricity generation of 34,000 MWh/year. It has been operating since 2011.

Credit: Green Prophet via Flickr

We covered the basics of light concentration by parabolic reflectors in Lesson 2. Now, we go on to look at all different aspects of the parabolic trough technology, including materials, operation parameters, system design, field layout, energy storage associated with this kind of plant. Please refer to the following reading material:

Based on the above reading, please go through the following self-check questions.

Unlike linear concentrating systems (troughs), which reflect light onto a focal line, the central receiver systems send concentrated light onto a remote central receiver. A typical example of such a system is a solar power tower system, which consists of multiple tracking mirrors (heliostats) positioned in the field around a main external receiver installed on a tower (Figure 7.2). Such systems are capable of reaching of much higher levels of concentration than linear systems. Concentrated radiation is further used as heat to produce steam and convert it to electricity (like in a regular power plant), or the generated thermal energy can be stored in a molten salt storage.

Figure 7.2. SOLUCAR PS10 (Planta Solar 10) solar power plant, Spain. Operational since 2007, PS10 produces 23,400 MWh/year. Aperture size is estimated at 74,880 m². The light is concentrated on the top of the 115 m high tower.

Credit: Afloresm via Wikimedia Commons

Please proceed further to the following chapter reading, which covers some basic configurations and design of the central receiver CSP technology, approaches used to convert and store energy, specifications of the heliostats and receivers utilized in some known facilities of this type.

It is clear that central receiver CSP systems are typically large-scale plants that are usually built to power a steam cycle. The central position of the receiver offers a universal advantage to collect all energy at one location and save on transport networks. At the same time, the fixed position of such a central receiver results in limitation of light collection: heliostats are always oriented at an angle to the direct beam, so the amount of energy collected is less compared to parabolic concentrators. Therefore, to reach the necessary efficiencies of light concentration, the size of the collecting field is increased, which brings into considerations such issues as land use, higher environmental impacts, and higher capital costs. Significant potential for developing large-scale central receiver solar plants is hence attributed to deserts and flat arid areas, which have plenty of sunshine and lower land value with respect to other applications and industries.

Parabolic dish geometry concentrates light in a single focal point, i.e., all sun rays that are parallel to the axis of the parabola are directed towards the central receiver. This allows this type of collector to achieve the highest concentration ratios among all other type of solar collectors. The dish concentrator must be oriented towards the sun. Usually, losses in this technology are associated with the imperfections of dish alignment and non-ideality of reflection. The engine that converts the concentrated solar energy into electricity is placed at the focal point. This technology can be used for both large-scale power plants (with many dishes grouped in arrays) and autonomous small-scale power generation systems that would provide power to off-grid remote facilities. Example of such a system is shown in Figure 7.3.

Figure 7.3. EuroDish stirling parabolic dish solar collector is an example of de-centralized solar technology. Some spec: diameter 8.5 m, aperture 56.7 m², average concentration factor 2500.

Credit: Schlaich Bergermann via Wikimedia Commons

Go to the following chapter reading to learn the fundamentals of the parabolic dish CSP technology:

As we can see from this reading, the parabolic dish is a very efficient and flexible technology, which comes in various designs and is suitable for various applications. There are many cases (some described in the history section) of applying this technology in remote locations to provide self-sustained power for water heating, water pumping, and alternative power. One of the features of the dish Stirling systems is their low "inertia" - they start producing power very quickly as soon as direct beam radiation hits the reflector, but, at the same time, they are very sensitive to variations in solar intensity and results in sharp variation in power output and frequent interruptions if the meteorological conditions are not perfect. Because of the high intermittence, energy storage applications which would buffer the power output, would be desirable, but at the moment are not well developed.

To make usable energy from solar heat collection in CSP plants, thermodynamic power conversion cycles (heat engines) are used. The main idea is quite simple. The heat transfer fluid, which is directly heated in the solar receivers, delivers heat to the boiler, which generates steam. Further steam is used in the heat engine to generate mechanical work to run electric generator. This scheme is very much the same as at conventional fossil fuel power plants, except for the heat being created not through combustion but through concentration of solar radiation.

General concept of solar thermal conversion system

Credit: Mark Fedkin (modified after Duffie and Beckman, 2013)

In general, the thermodynamic power cycles can be categorized into gas cycles and vapor cycles. In gas cycles, the working fluid is only present in the gas phase throughout the entire cycle. In vapor cycles, the working fluid can be transformed from the vapor phase to the liquid phase in different parts of the cycle. Rankine cycle, commonly used in stationary steam power plants, is an example of a vapor cycle with water as a working fluid.

In the course of a thermodynamic power cycle, the working fluid goes through a series of temperature and pressure (T,P) parameters. Changes in temperature and pressure result from heating, condensing, pressurizing, and expanding the fluid medium. Expansion creates the physical force to perform mechanical work, which is the main purpose of the system. The efficiency of the cycle is typically higher when the difference between the lowest and highest temperatures is maximized. For example, Carnot efficiency is:

$$\eta =\frac{T_{\max }-T_{\min }}{T_{\max }}$$

The maximum temperature is set by the heating source - for example, solar concentrator. The minimum temperature is set by the ambient conditions or cooling system - for example, air, river. The Carnot efficiency of 50% is considered good for a real system.

Several thermodynamic cycles that may be considered in connection to solar thermal applications are:

• Rankine Cycle
• Brayton Cycle
• Stirling Cycle
• Kalina Cycle

One of the criteria to combine these cycles with the solar thermal plants is the compatibility of temperature. The concentrating technologies must be efficient enough to generate high temperatures for efficient power conversion in the thermodynamic cycle. So, depending on the technology and type of solar collectors, one or another cycle may be chosen.

Another important criterion to consider is the choice of the working fluid. For higher temperatures (600 ^oC), steam is the best choice. For lower temperature conversions (100-400 ^oC), organic fluids are more suitable (Batton, 2000). The desirable properties of the working fluid are:

• low cost
• non-corrosive nature
• thermal stability
• high cycle and turbine efficiency

Possible choices for the working fluids include:

• water
• refrigerants
• organics
• ammonia
• toluene (under development)
• mixtures of above
• fully fluorinated benzene ring fluids

Currently, Rankine cycles are most promising for handling the collector temperatures.

We are going to overview the principle of thermodynamic cycle operation using Rankine cycle example, since most of solar power cycles currently operating are Rankine cycles.

The Rankine cycle system consists of a pump, boiler, turbine, and condenser. The pump delivers liquid water to the boiler. The boiler heated by the solar heat converts water to superheated steam. This steam is used to run the turbine which powers the generator. Steam leaves the turbine and becomes cooled to liquid state in the condenser. Then the liquid is pressurized by the pump and goes back to the boiler. And the cycle continues. Schematically, this process is represented in Figure 10.1.

Figure 10.1. Rankine cycle layout.

Credit: Andrew.Ainsworth via Wikimedia Commons

In an ideal Rankine cycle, all the units operate with the steady-state flow, and the kinetic and potential energy of the fluid are considered to be negligible compared to the temperature and pressure effects.

The work terms for each component of the cycle can be expressed as follows.

The work done by the pump to compress water (W_pump) can be represented as the change in enthalpy (H) of the water (fluid) before entering the pump and after leaving the pump:

$$W_{\text{pump}}=H_2-H_1$$

In this case, we assume there is no heat exchange with the surroundings, so the energy is not lost (which is an ideal scenario). The process, which is not accompanied by any heat exchange with the environment, is termed "adiabatic." So, this step 1-2 is adiabatic compression.

The next process 2-3 takes place in the boiler. The energy balance in the boiler can be expressed as the change in enthalpy of the fluid from the "before" state (compressed liquid) to "after" state (superheated steam):

$$Q_{\text{in}}=H_3-H_2$$

Where Q_in is the heat used by the boiler. This heat is supplied to the boiler from the solar concentrator. There is no pressure change in the boiler, only heat transfer to the fluid; therefore, no mechanical work is done here.

The next process 3-4 is expansion of the steam in the turbine. The work done through that process is the useful work, which is the main purpose of the cycle:

$$W_{\text{turbine}}=H_3-H_4$$

Here, we again assume that there is no heat exchange with the surrounding, so all the fluid energy change is converted to work. Note that the enthalpy change is written as "before" minus "after" because the energy of the superheated compressed steam is higher than the expanded steam after it exits the turbine. So, this expression gives us the positive work value.

Lastly, the process 4-1 is steam cooling and condensation. The energy balance on the condenser will be:

$$Q_{\text{out}}=H_4-H_1$$

At this stage, the extra heat is withdrawn from the system, and water returns to liquid state. This is important for the Rankine cycle from technological standpoint, since pumps employed in the system require a liquid medium to work efficiently and may have problems with water-vapor mixtures.

The energy balance for the whole cycle is then can be expressed via the following equation:

$$(Q_{\text{in}}-Q_{\text{out}})-(W_{\text{turbine}}-W_{\text{pump}})=0$$

The net work done by the system is W_turbine-W_pump. Therefore, the thermal efficiency of this cycle can be presented as follows:

$$\eta =\frac{W_{\text{net}}}{Q_{\text{in}}}=1-\frac{Q_{\text{out}}}{Q_{\text{in}}}$$

The basic Rankine cycle is presented in terms of temperature and entropy change in Figure 10.2. The ideal state of this cycle is reflected in the vertical lines 1-2 and 3-4, when the fluid compressed and expanded. Those processes are shown to proceed isentropically, i.e., without entropy change. That rarely happens in real life. Some fluid friction losses and dissipation of some heat to the surrounding usually makes this system deviate from the ideality (as for example, shown by the dashed lines).

In a non-ideal cycle, fluid friction results in the lower pressure in the line. To compensate for this pressure drop, the water needs to be pumped to a higher pressure. Heat loss can happen when steam flows through the connecting pipes and the cycle components, which are not perfectly insulated. To maintain the same work output, more heat needs to be transferred to the steam in the boiler.

Figure 10.2. Temperature-entropy diagram of the ideal Rankine cycle.

Credit: Olivier Cleynen via Wikimedia Commons

Varieties of the Rankine Cycle

There are several scenarios of employment of the Rankine steam cycle in power plants, including solar plants. Those scenarios intend to increase the overall efficiency of the system.

There are three ways to increase the efficiency of the basic Rankine cycle (Gramoll, 2015):

1. Decreasing condenser pressure. This results in lower heat rejection temperature of the fluid in the condenser (pushing point 4 on the diagram in Fig. 10.2 downward), thus allowing the system to produce greater net work.
2. Superheating steam to a higher temperature allows achieving higher temperature differential, thus increasing the amount of work done by the cycle.
3. Increasing the boiler pressure results in higher average steam temperature in the boiler. This effect allows additional work to be done in phase 2-3 (Fig. 10.3c). However, there is some loss of useful work in phase 3-4 because of necessity to re-heat the steam. Reheating is used to mitigate higher moisture content of the high-pressure steam.

The above-described efficiency modifications are illustrated in Figure 10.3.

Figure 10.3. Effects of different parameters on the work output of the Rankine cycle: (a) effect of decreasing condenser's pressure, (b) effect of superheating steam to a higher temperature, (c) effect of increasing boiler pressure. The shaded area shows the extra work performed by the system due to each parameter change. The red curve corresponds to the basic cycle, and green curve shows the adjusted cycle.

Credit: Gramoll, Kurt. Chapter 10 Rankin Cycle-Thermodynamics Theory. Archived December 4, 2021. Accessed May 21, 2025. Wayback Machine.  

Another variety of Rankine cycle is the Regenerative Rankine. The idea behind the regenerative cycle is to increase the temperature of the feed water that is supplied to the boiler. Why is it desirable? Higher water temperature would allow some energy savings for steam generation. So, some of the steam that exits the turbine is used for pre-heating the feed water for the boiler. This process is called regeneration. Heat transfer can be achieved using a heat exchanger (regenerator). There are two types of feed water heaters: open and closed.

The open feed water heater is essentially a mixing chamber, where the steam extracted from turbine is combined with water from the pump. The fluid that exits the mixing chamber is saturated water. The closed feed water heater is a heat exchanger, inside which the steam condenses on the outside of the tube carrying the feed water. As a result, the feed water temperature is increased. The pressures of the steam and the feed water do not need to be matched since the flows are not mixing. Without heat regeneration, the feed water temperature would be much lower and would require more energy from the heat source for the boiler. Regeneration helps raise the overall efficiency of the system.

Figure 10.4. T-S diagram of the regenerative Rankine cycle. In case of open feed water-heater, the phase 2-7 corresponds to the mixing, and 7-8 is the second compression of the fluid before the boiler.

Credit: Donebythesecondlaw via Wikimedia Commons

In this lesson, we looked at the types, design, and components of the Concentrating Solar Power (CSP) systems - parabolic trough, solar power tower, and parabolic dish. All of these types of CSP systems essentially differ in the way how they collect and concentrate light, and how the thermal energy is transferred. There is not as much difference in the final application (with some exceptions of course), as the output energy is primarily used to produce steam and run a turbine for electricity generation. The second half of this lesson reviewed the power converion cycles that accomplish that task. Based on this lesson, you should have a clear idea on the principle of operation of each of those types of systems and should be able to explain and compare the key specifications of CSP plants. Certainly, planning of a large-scale CSP plant is a complex task, which includes not only the technical principles of energy concentration and conversion. The major decisions are also made regarding the climate setting, solar resource, policy, and cost. Those considerations involved in strategic planning of CSP projects will be addressed further in Lesson 8.

The thermodynamic power conversion cycles used to convert heat into mechanical energy (and eventually electric power) are not novel technologies. Most of those cycles were technologically developed through 20th century and were widely applied in traditional fossil fuel fired plants. However, integration of those systems in solar CSP power plants may require some adjustments to match both system parameters. Rankine cycle has been most applicable to solar power plants to date as it operates in the lowest temperature range and therefore requires the lower concentration ratios. Stirling and Brayton cycles require higher temperature range and are suitable for higher concentration ratios, like those in central receiver systems. This lesson gives you a few examples of practical implementation of power conversion cycles in existing solar facilities.

Lesson 8 will cover a few important issues related to planning the large-scale CSP facilities. As we have already understood, performance of solar systems is very dependent on the locale, weather, and other physical conditions. Therefore, selection of a good site for a CSP project is a strategic step to take in order to ensure that the technology performs to its potential. So, criteria to consider in site selection are addressed in this lesson. Next, we will take a look at some factors in the social and economic spheres that pertinent to success of the solar projects. Finally, this lesson also touches upon the life cycle assessment of the CSP technology - LCA study examines a variety of metrics, which help to constrain the project with available resources and produced environmental impacts. In this lesson, you will be assigned to learn from a couple of published papers which present different methods of project assessment. Also, a lesson activity done in SAM software will give you some practical exercise with system parameters.

Learning Outcomes

By the end of this lesson, you should be able to:

1. understand the prerequisites for location of a CSP plant;
2. demonstrate the basics of the feasibility analysis for CSP systems;
3. define the key metrics used in socio-economic and environmental assessment of the CSP projects;
4. generate CSP system performance data in SAM software.

Solar energy systems rely on the natural solar resource, which is unevenly distributed over the planet. Clearly, some locations may be considered very favorable for harvesting sunlight, while others - much less favorable for a number of reasons. Choice of the CSP development site has therefore a strategic importance for the long-term feasibility of the project. Meteorological availability of sunlight is the first, but not the only, limitation to be considered when planning a CSP site development. One should also look at the physical geography of a potential site, available land area, available infrastructure, energy market, political and social situation. All these factors, being location-specific, can have a critical influence on the success of the CSP development in the area. In essence, the project development approach to CSP is no different from other project developments, such as those involving conventional fossil fuel power plants, photovoltaic solar, wind, etc. However, CSP has its own specific features that require special knowledge when site selection and feasibility study are performed.

Any site targeted for CSP should be carefully examined with respect to the key criteria. The most important of them are listed below:

• Direct normal irradiance (DNI). This is the amount of solar radiation received by a unit area that is perpendicular (i.e., normal) to the incoming sun rays. DNI converts to energy, which, in turn, converts to monetary return from the project. It is generally believed that DNI should be at least 2000 kWh/m² per year to provide a viable energy yield [Lovegrove and Stein, 2012]; however, the threshold would depend on the local market, and should be assessed individually for each case. The accuracy of the DNI data is very important for estimating predicted energy output from the future plant. It is recommended to perform on-site DNI measurements for at least one year to characterize a complete seasonal cycle.

  

  Figure 8.1. World map of DNI.

  Credit: SolarGIS © 2014 GeoModel Solar. Refer to the SolarGIS website to get the DNI maps for specific locations.
• Available land area and topography. These conditions are different for different CSP technologies. For example, parabolic trough setup tolerates up to 2^o slope, while Fresnel reflector systems can tolerate up to 5^o slope. Solar tower systems can accommodate more variegated topography as long as the slopes properly support the arrangement of the heliostats. Typically, utility scale CSP plants requires substantial area of open land (~ 2-5 km²) free from obstructions. Typical land use by CSP plants was estimated by NREL at ~40,000 m²/MW [NREL, 2013].
• Soil and subsoil structures. The land should allow some leveling works to be performed. This requires the analysis of soil and rock conditions. Some soft soils may be unstable and undermine the alignment of optics; so, it may need to be replaced. Rocky formations underneath may bring extra costs for leveling works.
• Weather profile. Long-term data analysis of meteorological and satellite data helps to create predictions for the cloudiness in the area over the system lifetime. Also, strong tendencies for rain, water run-off, or flooding may have effect on site topography. Strong winds may have a direct impact on power generation since they create vibrations and deformations on concentrators and thus affect sunbeam focusing precision. Some sites benefit from wind-breaking barriers and other mitigation measures. Dry bulb and wet bulb temperatures are assessed to define the achievable cold-end temperatures of the power plant. This has effect on the steam turbine efficiency.
• Local water resources. Water is needed for CSP plants first of all as the cooling agent for steam turbine condenser. Both subsurface and surface water can be used. However, the locations with highest DNI are typically arid areas (e.g., deserts), where water is scarce. Associated need for water use authorization may become an additional obstacle for site approval. Other water use includes service, mirror cleaning. Water-steam cycle requires demineralized water, which would require an additional on-site water treatment plant.
• Grid connections. The power produced must be delivered to customers, so the connection to the high voltage electricity line is necessary. CSP plants of 20 MW and above capacity usually use the lines in the range of 60-400 kV. The distance to the grid should be minimal to avoid additional investments for power transmission. If auxiliary routes for power transmission need to be constructed, authorization and environmental concerns should be addressed up front.
• Proximity of roads. Roads are needed at the stage of plant construction for transport of materials and equipment to the site. Further, at the operation stage, access roads become permanent structures.

The simultaneous analysis of the multiple factors make the site assessment a complex iterative process. The main successive phases of this process are illustrated below (Lovegrove and Stein, 2012):

1. Market Analysis

2. Regional Stud and Site Identification

3. Feasibility Analysis

4. Project Qualification

5. Contract Closing

Each subsequent stage in the above scheme requires more specific information and additional expertise. This process of site characterization and selection is to some extent illustrated in the report by Stoddard et al. (2005) - the study that considered various alternatives for CSP plant development in New Mexico:

Water considerations

The water use strategy at the CSP facility is one of the keystones of the project, as it would influence many other factors and choices and may become a go/no-go tipping point for the project. This is because water shortage is being identified as a severe environmental problem in many regions, and, therefore, is subject to strict environmental regulations. So, let us take a closer look at the water requirements in CSP systems and associated technological options.

The biggest consumption point of water in CSP is cooling for the steam turbine condenser. Cooling systems can utilize both salt and fresh water, which can be taken from both surface and subsurface reservoirs. Agreement on water withdrawal should be reached prior to the project start, and that should be part of the project feasibility analysis. There are wet-cooling and dry-cooling systems. The wet cooling usually implies using a cooling tower, and that is the most efficient technology for cooling as long as water is affordable and continuously available for plant operation.

If you are not familiar with those, watch the following video (3:16) to see how they work:

Video: How Cooling Towers Work (3:16)

However, cooling towers use up to 85-90% of all process water. When water is in short supply, dry-cooling systems can be an alternative. Dry-cooling systems use ~10 times less water than wet cooling systems do; even truck-based supply may suffice. However, from an economic point of view, dry cooling systems are less beneficial. In dry cooling systems, air is used as heat transfer medium, and air has much lower heat transfer coefficient than water. Furthermore, the cooling effect of evaporation, which is the core mechanism of cooling in cooling towers, is not available. This results in lower efficiency of the water-steam cycle. Another drawback of the dry-cooling systems is additional power consumption by fans blowing air for cooling. For the above reasons, the wet-cooled projects have an economic advantage over dry-cooled projects. The decision involving the trade-off of water versus energy is to be made individually in each particular case based on available resources. One of the compromise options for water use is hybrid cooling tower, which combines dry cooling and wet cooling. In this technology, water is sprayed on the condenser allowing for evaporative action, but the water consumption is significantly lower compared to conventional wet cooling method.

According to water use estimation by Andrew Eilbert (Worldwatch Institute), on the average, CSP plants use only 120 Gal of water per megawatt-hour of energy. For comparison, this number is lined up against typical values for other types of power plants in Table 8.1. [Eilbert, 2010]. Visit the WorldWatch website for specific data on water use by different CSP plants in California.

Other water uses in CSP plants include:

• process (water-steam cycle)
• service (cooling rotating equipment)
• washing (mirror cleaning)

The water involved in water-steam cycle needs to be relatively pure and often needs to be de-mineralized. Requirements for water purity is specified by the turbine manufacturers. These requirements impose an additional limitation on the water sources. If raw water contains significant amounts of ions and other chemicals, a special water treatment plant may need to be added to the facility. Much of this water is recycled, and its total volume is not substantial.

Please answer the following self-check questions before proceeding to the next section.

CSP deployment has a number of positive collateral impacts on environment and social welfare, and those impacts are important to consider in the project feasibility analysis. Sometimes, when project evaluation is solely based on the energy prices, renewable energy technologies may not look competitive enough at the modern energy market. However, including externalities, such as greenhouse gas emission reduction effect, improved diversity, security of the energy supply, employment, etc., into the evaluation process can help to justify the value of a renewable energy project more fairly.

One of the prominent impacts of CSP in the socio-economic area is stimulation of the economy and creation of new jobs at the local level. This is largely due to relatively "low-tech" profile of this technology: the main components are mirrors, steel, concrete, and labor. These local impacts are realized through the increase in demand for goods and services and creation of jobs. These impacts can be classified as direct, indirect, and induced.

The direct effects imply the increased demand for good and services that are required to construct, operate, and maintain the CSP facility. The indirect effects involve the one the new investment has on new sales and material flows among other productive sectors of the economy. The induced effects are related to expansion of private expenditure (for example, from workers employed) in goods and services, such as food, health care, transportation, etc. [Lovegrove, Stein, 2012]. For proper accounting of all levels of external benefits, there should be a way to quantify them and to assign them a monetary value.

One of the analytical methods to quantify the externalities of energy projects is based on Input-Output (I-O) analysis. The I-O analysis is an economic theory that was developed by Wassily Leontieff, a Russian economist who received a Nobel Prize for it. This model describes the inter-industry relationships within an economy, connecting the outputs from one part of the economy to the inputs to another part of the economy. The data are typically expressed as monetary values and are organized as a matrix, with column entries representing inputs to an industrial sector, and rows representing outputs from that sector. [Input-Output Model from Wikipedia.org ]. Essentially, this approach recognizes that spent investment becomes income to other industries, and thus stimulates their development.

To understand how this method can be applied to CSP projects, we refer to the following publication, which analyzes the socio-economic impacts of parabolic-trough and solar-tower plants in Spain. The authors come up with impressive numbers for increase in demand and employment impacts, demonstrating the remarkable potential of CSP for benefiting the local economy.

This article nicely demonstrates the quantified benefits from solar thermal energy projects, which are rarely taken into account in feasibility analysis. The type of analysis presented there is a convenient tool to determine whether the government subsidies provided to defray the cost of the renewable technology are justified in terms of social welfare.

Some main points in the above-mentioned study are included in the self-check questions below.

As a technology based on the renewable solar resource, the CSP has a great potential to reduce greenhouse gas (GHG) emissions. Emissions typically associated with conventional electricity production result from coal or natural gas burning and have been a severe factor in global pollution in climate change. Hence, installation of CSP offsets generation from fossil fuel plants. It does not mean, however, that CSP technology is emission free: while operational emissions are negligible compared to fossil-fuel power plants, lifecycle emissions may be still significant. To estimate the GHG emission level and the magnitude of the emission reduction benefit, one can employ Life Cycle Assessment (LCA) - a comprehensive methodology dealing with inventory of all processes and materials involved in a technology.

Life Cycle Assessment (LCA) methodology

Life Cycle Assessment (LCA) is a “cradle-to-grave” approach for assessing products, processes, industrial systems, and the like. “Cradle-to-grave” begins with the gathering of raw materials from the earth to create the product and ends at the point when all materials are returned to the earth. LCA evaluates all stages of a product's life from the perspective that they are interdependent, meaning that one operation leads to the next. LCA enables the estimation of the cumulative environmental impacts resulting from all stages in the product life cycle, and, as a result, allows selecting a path or process that is more environmentally benign.

LCA helps decision-makers select the product, process, or technology that results in the least impact to the environment. This information can be used with other factors, such as cost and performance data, to find optimal solutions. LCA identifies the transfer of environmental impacts from one media to another (for instance: a new process may lower air emissions, but creates more wastewater, etc.) and between different lifecycle stages. The diagram below illustrates the main lifecycle stages to be considered in LCA:

Figure 8.2. The main flows and stages considered in lifecycle assessment.

Credit: Mark Fedkin

The diagram above lists the main stages of product lifecycle: (1) raw material acquisition, (2) manufacturing / construction, (3) operation or use, and (4) recycling or waste disposal (decommission stage). Each of these stages has inputs of materials and energy and outputs of atmospheric emissions, waterborne wastes, and solid wastes. Each stage creates the main useful input to the next stage, and usually the operational stage of the lifecycle is where the main product of the technology is produced. Any co-products (desirable or undesirable) are also identified and taken into account in the analysis. The LCA based on this scheme is a complex process (even for small systems), which requires large amount of data and interdisciplinary expertise for proper assessment. A typical LCA project plan includes the following main steps:

1. Goal definition and scope: Identify a product / process / technology; establish context and system boundaries.
2. Inventory analysis: Identify and quantify energy, water, and materials as inputs as well as environmental releases as outputs.
3. Impact assessment: Assess the potential human and ecological effects, quantify metrics.
4. Data interpretation: Compare data from Inventory Analysis and Impact Assessment stages to select or recommend a preferred product, process, or technology.

LCA Limitations:

• LCA thoroughness and accuracy will depend on the availability of data; gathering of data can be problematic; hence, a clear understanding of the uncertainty and assumptions is important.
• Classic LCA will not determine which product, process, or technology is the most cost-effective or top-performing; therefore, LCA needs to be combined with cost analysis, technical evaluation, and social metrics for comprehensive sustainability analysis.
• Unlike traditional risk assessment, LCA does not necessarily attempt to quantify any specific actual impacts. While seeking to establish a linkage between a system and potential impacts, LCA models are suitable for relative comparisons, but may be not sufficient for absolute predictions of risks.
• The standardized procedure for the LCA recommended for product and technology assessment in the U.S. is documented in the EPA guidelines referenced below.

Life Cycle Assessment: Principles and Practice, EPA/600/R-06/060, 2006.

LCA Methodology Applied to CSP Plants

When the LCA analysis is applied to a CSP plant, a number of impact metrics need to be identified. One of them, as mentioned above, is greenhouse gas emissions, but there are also other impacts that involve environmental harm. Other typically assessed impacts are acidification and eutrophication potential. Acidification is referred to as increase in acidity of natural waters and soils. It may cause loss of aquatic life, forests, and other plants in the area and increase ecotoxicity. Eutrophication is related to nutrient enriching in aquatic and terrestrial environments. It can cause harm to ecological system through a chain of feedbacks, change in dissolved oxygen contents.

With respect to above metrics, CSP offers significant benefits, according to various studies reviewed in Lovegrove and Stein (2012). Values of GHG emissions by CSP plant (in case of solar only operation - no hybrid systems with natural gas backup) are estimated in the range from 11 to 90 g CO2 equivalent per kWh of electric energy generated. For central tower plants, the emissions are on the lower side, and for the parabolic trough plants, the emissions are on the higher side of that range; but, in either case, these numbers are well below the values typical for other electricity generation facilities (Table 8.2).

The CSP emissions have been shown to increase by 650 g CO₂ equiv./kWh if a fossil-fuel back-up is used and by 60 g CO₂ equiv./kWh if heat storage is used. The different studies indicate that CSP's GHG emissions are mainly associated with steel and concrete used in solar field and tower, and salts used in storage systems. The use of synthetic salts instead of naturally mined salts for storage system results in an increase in GHG emissions by ~13 g CO₂ equiv/kWh. Also, non-renewable electricity and materials may be used at the manufacturing stage of the project and in transportation to the site. It is known that the use of dry-cooling system instead of wet-cooling system increases GHG emissions by ~2 g CO₂ equiv/kWh.

Acidification impact of CSP reported in the range 70-100 mg SO₂ equiv/kWh. In case of hybrid plant (solar + fossil fuel backup), this number is substantially higher - 590-612 mg SO₂ equiv/kWh. In case of eutrophication, the impact range is 6-10 mg PO₄/kWh for solar only mode and ~50 PO₄/kWh for hybrid mode.

It is important to note that these impacts take place during the operational phase only because the plant consumes power from the grid rather than power being produced on site.

A published study by Lechon and co-authors (2008), referenced below, describes the LCA analysis for CSP systems in more detail. The first two sections of the article define the method and scope of LCA, so you can just briefly look through those. It is more important to look at the data tables that contain specifications of the assessed facilities and various metrics. Usually, LCA studies collect considerable amounts of data, and our goal here is to learn to read and to interpret that information.

As you can understand from the above article, the energy consumption at different stages of the system lifecycle is still the main source of the environmental impact. The numbers characterizing two different CSP facilities in terms of energy consumed and energy generated are listed in Table 3, which is the basis for finding the lifecycle emissions. Also, pay attention to Table 6, where the emissions are itemized by system component. This itemization can be helpful in finding engineering solutions to further decrease the environmental impact of the technology. Note that the plants assessed in this study are hybrid (solar thermal + natural gas 15%), and this is the reason for higher overall GHG emissions calculated for this case compared to solar-only values given in Table 8.2 above.

The biggest challenge in prediction of the CSP performance coming from a projected solar thermal plant is the unsteady nature of the solar resource. Models involve weather and insolation data which bring in uncertainties in the final power output. So, it is not sufficient to calculate the annual energy yield simply by expected load hours, as it is usually done for conventional fossil fuel power stations.

The energy flow within a CSP plant follows a complex chain of transformations, each provided by a certain technology block. This step-wise energy flow can be represented by the following diagram:

Figure 8.3. Successive transformation of natural power to electric power in the CSP system.

Credit: Mark Fedkin

Each step in the power conversion chain is performed by technological units, each characterized by a number of parameters, which can be optimized depending on the scale of the facility, external conditions, available resources, and other preferences. Simultaneous optimization of those parameters and computation of the power output can be done by System Advisor Model (SAM) software, distributed by the National Renewable Energy Laboratory (NREL). SAM currently is one of the widely used and most developed tools for solar plant modeling, which incorporates both energetic and economic parameters. SAM has a user-friendly graphic interface, which allows the operator to see diverse technical performance information in table and graphs.

In SAM, the setting for each of the above mentioned technological units can be predefined and controlled separately, which creates opportunities for simulating different solar systems at different locations.

If you are not yet familiar with how the SAM software works, please refer to the following video (35:08), which explains the basic things about the program interface.

Video: SAM Intro Webinar 1 of 5: User Interface Overview (35:08)

As an activity in this lesson, you will use a few examples from SAM to explore the differences between different cases of CSP installations.

Different types of fluids are commonly used for storing thermal energy from concentrating solar power (CSP) facilities. CSP plants typically use two types of fluids: (1) heat-transfer fluid to transfer the thermal energy from the solar collectors through the pipes to the steam generator or storage, and (2) storage media fluid to store the thermal energy for a certain period of time before it is used on demand.

These are some available heat transfer and storage fluids currently used:

• Water
• Water-Glycol mixtures (for low temperature only)
• Mineral oils
• Synthetic thermal fluids
• Molten inorganic salts
• Molten metals

Water is the most available and cheapest fluid to use, but the problem with water is that it has very limited temperature range when it is liquid. Keeping water in liquid state above 100 ^oC requires high pressure, which adds significantly to system complexity and cost. Oils and other synthetic liquids are commonly used in CSP plants, as they have a much wider working temperature range. Molten salts are probably the most common storage medium (Wu et al., 2001), but are not the best heat transfer medium, because salt tends to solidify in tubes at lower temperatures, blocking transport. Additional heating will need to be provided in that case to start up the plant. There are developments for novel fluids that can be used for both heat transfer and storage at the same time (Moens and Blake, 2004). Some commonly used fluids and their working temperature ranges are shown in the diagram in Figure 9.1.

Figure 9.1. Different types of heat transfer and heat storage fluids and their approximate working temperature ranges.

Click here for a text description of Figure 9.1

Mineral oils: -50^oC to 310^oC, Synthetic thermal fluids: -50^oC to 340^oC, Molten inorganic salts: 142^oC to 500^oC, Molten metals: 97^oC to 650^oC, Water: 0-100 Water pressurized up to ~ 350^oC

Credit: Mark Fedkin, modified from Koning, 2007.

Molten salt storage is employed at many existing solar thermal plants, so we are going to look at it in some more detail. As was mentioned above, salt "freezing" (i.e., transforming from a molten state to solid state) can present some problems, because the salt is supposed to circulate through the tubes to deliver thermal energy to the steam generator or another application. So, to achieve the lowest possible "freezing" temperature, a eutectic mixture of salts is used.

What is eutectic?

A eutectic system is a homogeneous mixture of two or more components, which together have a lower melting point than each of them separately. The eutectic mixture melts as a whole only at a specific ratio of those two components in the mixture. A generic eutectic phase diagram is shown below.

Figure 9.2. Generic phase diagram of an eutectic system. The horizontal axis measures the composition of mixture of components of A and B - point A represents 100% of component A, and point B represents 100% of component B. Any composition in between is the proportional mixture. The vertical axis is temperature. The melting temperatures of pure A and B are marked on left- and right-hand sides respectively. The eutectic mixture of A and B has much lower melting point (marked by the red dot).

Credit: Mark Fedkin.

Typical molten salt mixture used for energy storage is represented by the ternary eutectic

53 wt % KNO₃ (potassium nitrate)
40 wt % NaNO₂ (sodium nitrite)
7 wt % NaNO₃ (sodium nitrate)

or binary eutectic

45.5 wt % KNO₃ (potassium nitrate)
54.5 wt % NaNO₂ (sodium nitrite)

The eutectic temperature for those compositions is in the range 142 to 145 ^oC.

The molten salt is used for high temperature energy storage applications (above 400 ^oC) because typical thermal fluids, such as synthetic oils, have temperature limitation and decompose beyond the maximum allowable temperature (410-430 ^oC).

This is how the molten salt storage is employed in a solar thermal plant. First, the solar energy is caught by the collectors and concentrated on the receiver tube filled with heat transfer fluid. The heat transfer fluid (with temperature of ~393 ^oC) is circulated in a closed loop to deliver heat to the steam generator, which produces superheated steam, and then the thermal fluid flows back to the solar collectors (with temperature of ~ 293 ^oC). Such a loop can only operate during sunshine hours. To extend the steam generation beyond sunshine hours, molten salt thermal energy storage is used. The thermal storage usually consists of two salt storage tanks. In this case, the closed loop with the heat transfer fluid is passed through one of the salt tanks, where salt is heated to the temperature of ~ 384^oC. The tank is insulated, so salt can stay hot for a substantial period of time (estimated heat loss ~0.5 ^oC per day). The molten salt is stored in the tanks at ambient (atmospheric pressure). To discharge heat during night hours, the molten salt from the hot tank is pumped through the steam generator to produce steam, and then to the cold storage tank (at ~ 292 ^oC). In this configuration, some part of the heat transfer fluid loop is diverted to the heat exchanger between the cold and hot salt tanks. The cooled molten salt is then pumped through the heat exchangers and returns to the hot salt tank.

Solar tower systems can use molten salt as heat transfer fluid and heat storage medium without involving any additional thermal transfer fluid loops due to higher radiation concentration temperatures. In this case, molten salt is flowing through the tower-mounted molten salt receiver, where it is heated to 565 ^oC. Then the salt is supplied to the hot salt tank, from where it flows to the steam generator. This concept is illustrated on the eSolar website.

In this case, the use of molten salt for both heat transfer and thermal energy storage minimizes the number of storage tanks and salt volumes needed.

The following video (~2 min) provides a simple illustration of the molten salt thermal energy storage concept.

Video: Molten Salt Energy Storage (2:43) (on-screen text and animation set to music)

Storing energy in fluids involves exchanging heat between different types of fluids in heat exchangers. For example, transfer of heat from a thermal fluid in solar-heated tubes to the molten salt reservoir requires a heat exchanger; further transfer of heat from the molten salt to water to produce steam would involve another heat exchanger. There is a thermal physics method to calculate the effectiveness of the heat transfer in different types of heat exchangers and to evaluate their performance. One example of such calculation is given in Section 3.17. of the book "Solar Engineering of Thermal Processes" by Duffie and Beckman (2013), referred to below.

Strategic planning of a CSP facility involves a number of steps - site evaluation, socio-economic assessment, environmental assessment, and system modeling and optimization. All these steps must be taken before investment is made, and work has begun. Feasibility analysis requires several iterations and special expertise to justify the decision. This lesson introduced you to a few useful tools and methods to start with.

After you have covered the assigned readings for this lesson, please complete the following assignments:

This lesson will overview energy storage options for large-scale solar facilities. Clearly, energy demand rarely coincides with energy generation. Being bound to daily solar activity cycle at a certain locale, solar energy conversion systems are intermittent by nature, therefore, using energy at nighttime requires technologies to store energy on site. Photovoltaic systems, which convert natural solar resource into electric power, require means for electrical energy storage, while CSP systems may be better off storing thermal energy. The thermal storage principles and technology were discussed in the previous lesson, and Lesson 9 is primarily concerned with the technologies used to store electric power. For storage, electrical energy is often converted to other kinds of energy; for instance, potential mechanical, kinetic mechanical or chemical energy, which would be stored as fuel. Energy storage research has been accelerated over the recent years to address the need for compact and economically efficient storage technologies and is currently define the rate implementation of the commercial renewable energy systems. This lesson readings provide an overview and resources for learning the storage technology principles, with understanding that some of those options are much more advanced and rapidly evolving than others.  

Learning Outcomes

By the end of this lesson, you should be able to:

1. explain the technical principles of energy storage used in utility-scale solar plants;
2. understand the performance metrics for storage systems;
3. articulate technological, environmental, and economic cons and pros of energy storage implementation.

Because solar energy supply is variable in time, energy storage is an important issue. Energy storage is used to collect the energy generated by the solar conversion systems (thermal or photovoltaic) in order to release it later on demand. This can be a situation when sufficient power is produced during the day, and stored energy is used during the night. Also, when insolation conditions are ideal, the solar system may produce enough power for the target application, but on dull days, direct energy supply from collectors is diminished, and the energy from the storage is used to compensate the deficit. Energy storage devices help to smooth out differences and minor fluctuations in energy supply caused by shading, passing clouds, etc. Development of efficient and cost-effective energy storage is considered the main bottleneck of the universal development of solar systems.

Video: How solar energy got so cheap, and why it's not everywhere (yet) (7:53)

There are quite a few different technology options for energy storage, which are briefly outlined below:

1. Grid. For grid connected solar systems, the most natural and cost-effective way would be to store energy in the grid. The main idea here is that the DC power from a solar facility (array or farm) is converted to AC power and is fed to the grid and further on is used for on-site or off-site applications. This way, the grid acts as the medium that collects energy from different power-making facilities (renewable or non-renewable) and redistributes it as necessary. Since a grid does not really represent a separate system that is part of a solar plant, it will not be discussed further in this lesson.
2. Fluid. Fluid-based storage is typically used with solar thermal systems. Unlike grid, which stores electrical energy, fluids store thermal energy. Fluids, such as water, oil, molten salt or others act as a medium for absorbing heat. The main idea is that the solar radiation heats the heat-transfer fluid which is accumulated in the tank. The tank is insulated, so the hot fluid keeps its temperature for a substantial period of time. When needed, the heated fluid is used in a heat-exchanger to produce steam for the electric generator. This type of thermal energy storage was discussed in more detail in Lesson 8.
3. Battery. A battery is an electrochemical device that stores chemical energy in internal components and releases energy as electricity, which is generated through electrochemical reactions. Batteries are reversible, i.e., can be charged and discharged, and the parameters of these processes are regulated to avoid damage by overcharging or over-discharging. Battery life is expressed in number of charge-discharge cycles. There are many different types of batteries, some of which will be discussed further in this lesson.
4. Hydrogen. The idea behind hydrogen storage is that electricity generated by solar PV systems can be used to electrolyze water - to split it to hydrogen and oxygen. Further, hydrogen gas is collected and can be used as a fuel. One of the highly efficient devices "converting" hydrogen back to electricity is H₂/O₂ fuel cell, which has zero carbon footprint during operation.
5. Compressed air. In this case, the electrical energy produced by a PV solar system is used to run compressors to compress massive amounts of air and store it in underground, above-ground, or underwater containers. Later on, when energy is needed, the air is decompressed and is supplied to a turbine to generate electricity. Compressed air energy systems (CAES) promise high efficiencies, although this technology is not yet widely implemented.
6. Pumped storage hydropower. The available energy can be used to pump water into an elevated reservoir for storage. When power is needed, the water can be discharged under gravity to run a turbine, which is connected to a generator to produce electricity. The same as compressed air systems, the pumped storage technology has high energy return on investment, although it may require special topographical conditions and water availability in order to be used.

All of the above options for energy storage should be employed with understanding the facility needs and capacity. What energy storage is efficient for small residential systems may be insufficient or too costly when scaled up to the utility-size systems. Determining capacity of energy storage for a particular solar project is an important technical and economic issue. For example, if the capacity of the storage is too large compared to the energy produced by the solar conversion facility, the total system cost will be unnecessarily increased. On the contrary, if the capacity of the storage is too small, that leads to energy dumping and overall unsatisfactory plant performance.

In the following sections, we will discuss different energy storage options that can be possibly applied to utility scale solar systems.

Batteries are commonly used to store electric energy generated by off-grid renewable energy systems, and also to mitigate the sharp fluctuations of power for on-grid systems. While there are many different types of battery technologies, some are more applicable to utility scale energy storage than others. Applicability to large systems depends on such factors as cost of materials, ability to scale up with no ill effects or performance loss, and design and operation mode.

Some well-known examples of battery types used as stationary storage system for PV solar are listed in Table 9.1

Note: The costs in the table are based on standard assumptions for the applications and technologies considered, and on expert opinion. They are meant to be used for comparative purposes. The actual costs of any storage system depend on many factors and the assumptions and the means of calculating some of the values are subjective and continue to be debated, even among experts in the field (Sandia National Laboratories).

For quite a while, lead-acid batteries have been the first choice for off-grid PV applications. This lead-acid battery technology has been around since the 19th century and, historically, service providers have more knowledge and tools to deal with those systems. But, despite their long existence and widespread use, lead-acid batteries remain one of the lowest energy-to-weight and energy-to-volume battery designs, which means they are too big and heavy for the amount of energy they provide. This technology is inexpensive and reliable, and it may be a while before it is replaced by more advanced types on a wide scale.

The following reading provides more information on the battery storage types and lead-based batteries, specifically.

Li-ion battery technology

Li-ion battery is one of the rapidly advancing technologies preferred for employment in conjunction with solar systems due to high storage capacity, high charging rates, light weight, and relatively long service life. However, the technology cost is still high and can be a limitation on the utility scale. Some of the very attractive features of Li-ion batteries are high power output and high charge-discharge efficiency. They can also withstand more charge-discharge cycles than lead-acid batteries.

The principle of operation of the Li-ion battery is discussed below.

A schematic representation of a generic Li-ion battery is given in Figure 9.1. Roughly, Li-ion cell consists of three layers: electrode 1 (cathode) plate (usually lithium cobalt oxide), electrode 2 (anode) plate (usually carbon), and a separator. The electrodes inside the battery are submerged in an electrolyte, which provides for Li+ ion transfer between the anode and cathode. The electrolyte is typically a lithium salt in an organic solvent.

Figure 9.1. Li-ion battery system and charge transfer processes

Credit: Wikimedia Commons / Mark Fedkin.

During the charging process, a DC current is used to withdraw Li⁺ ions from the cathode and to partially oxidize the cathode compound:

LiCoO₂ → Li_1-xCoO₂ + xLi⁺ + xe^–

The released Li⁺ ions migrate through electrolyte towards the anode, where they become absorbed in the porous carbon structure:

xLi⁺ + xe^– + C₆ → Li_xC₆

At the same time, electrons travel through the external circuit (electrolyte is not electron conductive).

During the battery discharge, the reverse process takes place. Li⁺ ions spontaneously return to the cathode, where electrochemical reduction occurs.

Limitations of the Li-ion batteries are rooted in the material properties.

For example, the LiCoO₂ ⇔ Li_1-xCoO₂ $$ conversion is only reversible with x<0.5, which limits the depth of the charge-discharge cycle. But with a wider variety of materials available, research is underway to develop new generations of Li-ion batteries.

For example, take a look at the Sigma Aldrich website, which lists multiple alternatives for cathode, anode, electrolyte, and solvents.

Flow Batteries

Flow batteries, unlike solid-state batteries, have their chemical components dissolved in liquid solutions, which can be pumped through the electrodes in a flow. If you are familiar with the concept of fuel cell, it is something similar in principle of operation, although it is still a closed loop system. A flow battery cell itself can be small, while the solutions can be contained in external storages. One of the advantages of the flow batteries is almost instant replacement of the electrolyte liquid, thus eliminating any gradient or concentration fluctuations at the electrodes. The main difference between the conventional batteries and flow batteries is that the energy is typically stored in the liquid phase in flow batteries. So, increasing the size of the storage tanks for the liquids allows easy scale-up of the battery to match a specific application.

Zinc-bromine flow battery storage

Zinc-bromine battery is a type of hybrid flow battery. It uses zinc bromine as the working solution, which is stored in two compartments, separated by a porous membrane. One compartment has a negative zinc electrode and the other compartment has a positive bromide electrode. During charge, supplied electricity (e.g., from a solar conversion system) is used to electroplate metallic zinc (Zn) on the negative electrode, while bromine (Br₂) is generated on the positive electrode. During discharge, the opposite process occurs: Zn is dissolved to form Zn²⁺ ions in solutions, and bromine is converted back to bromide ions (Br^-).

Here are the electrochemical reactions involved in this process:

Zn²⁺ + 2e^- → Zn(s)   - Reduction of zinc during battery charging

2Br^- → Br₂(aq) + 2e^-    - Oxidation of bromine during battery charging

The overall reaction is therefore:

Zn²⁺ + 2Br^-⇔ Zn(s) + Br₂(aq)

This reaction proceeds to the right on charging and to the left on discharging. The standard electrode potential for the overall reaction is 1.85 V, which is the maximum theoretical voltage that can be expected from a single cell. The battery cells are stacked to increase the overall storage capacity of the system.

The battery compartments are made of inert plastic. Unlike common batteries, which store electrolyte within the reaction chamber, zinc-bromine batteries have solution storage in the external tanks, from where it is circulated through the electrodes (flow battery type). The external bromide solution storage also helps maintain the required concentration of bromide throughout the reaction cycle.

This technology has been commercialized by ZBB EnerStore company, which engineered zinc-bromine batteries into 50 kWh modules, scalable up to bigger storage systems. Each module is a stand-alone system that includes all necessary software and hardware. Some advantages of this technology include high-energy density (75-85 Wh/kg), stability, i.e., good resistance to performance degradation, ability to operate at full output within a wide temperature range. Unlike most batteries, ZBB EnerStore batteries use non-reacting electrodes (i.e., electrodes are not reactants, but simply are substrates for reactions to take place), which helps minimize loss of performance from repeated cycling.

Watch this video (5:19 minute) for a demo of ZBB EnerStore solution for zinc bromide battery technology:

Video: ZBB EnerStore (5:19)

Vanadium Redox Flow Batteries

This type of battery utilizes the multiple redox states of vanadium (V) in its charge-discharge cycles. Vanadium is present in the dissolved form in the sulfuric acid medium, and because it is all-vanadium system, this type of battery is not susceptible to performance loss due to cross contamination.

During charging, the following half-reactions occur in two separate compartments of the battery:

V³⁺ + e– → V²⁺

VO²⁺ + H₂O → VO₂⁺ +2H⁺ + e^–

Electrons are supplied from the solar energy conversion system as DC current onto non-reacting electrode dipped in the V³⁺ solution. As a result, V³⁺ is reduced to V²⁺. At the same time in the other compartment, vanadium (IV) species VO²⁺ is oxidized to vanadium (V) species VO₂⁺, releasing the electron. On discharging, these reactions are reversed.

The summary process is expressed through the following reaction:

VO²⁺ + V³⁺ + H₂O ⇔ V²⁺ + VO₂⁺ + 2H⁺

The total voltage generated by a single vanadium redox flow battery is around 1.25 V in ideal case.

The main benefits of the vanadium redox flow batteries ability to go through "unlimited" number of cycles; they have a long lifespan (>20 years), quick charging, and high efficiency of the charge-discharge cycle (~80%). They are also more environmentally friendly in terms of component toxicity than many other types of batteries.

The vanadium redox flow battery technology is potentially suitable for extra-large utility scale applications. For example, the 200 MW VRB battery facility in Dalian, China, is expected to significantly increase the stability of the electric grid by supplying power during peak hours and emergency black-starts. Development of such a mega facility was enabled by its co-location with the VFB cell manufacturing factory, which is tapping into local vanadium resources. The Dalian battery is expected to become operational in 2020. Nearby wind power facilities have been forced to curtail electricity production – this battery facility hopes to reduce curtailing significantly.

Check out the story here.

Additional resources:

Review paper: Blanc, C. and Rufer, A., Understanding the Vanadium Redox Flow Batteries, in Paths to Sustainable Energy, Dr. Artie Ng (Ed.), ISBN: 978-953-307-401-6. pp. 333-337(Access the article in Canvas).

This paper is quite technical as it describes different models used to analyze the performance of the vanadium redox flow batteries. Read sections 1 and 2, which describe the electrochemical principles behind the battery operation. Reading further sections may be useful if you have a special interest in this topic, but is not required.

Compressed Air Storage

Compressed air storage technology may become an efficient solution of storing energy generated by large solar plants. The concept is as follows.

Air is used as the energy transfer medium. During the daytime, solar power is used to heat and compress air in an airtight chamber. When energy is needed, that compressed air can be expanded through a turbine or another expansion device to drive a generator to create electricity. Compressed Air Energy Systems (CAES) have been in use in some conventional power plants, and they are making a come-back as energy storage systems for renewable energy plants.

Traditionally, CAES technology used underground geological formations, such as salt caverns, as reservoirs for compressed air. While this approach was effective at some locations, it was not universal, as geology in some areas may be just unsuitable. A newer approach with CAES is to use human-made chambers - large pipes, such as those used for natural gas pipelines. While it involves more construction and installation, this type of artificial storage can be employed virtually anywhere and scaled up to the required capacity by simply using longer pipes. 

Based on this reading, answer the following self-check questions:

Pumped Hydro Energy Storage

Pumped-storage hydropower (PSH) is the type of storage technology that is based on storing energy in the form of potential energy of water. It consists of two water reservoirs at placed at different elevations connected by discharge channel. The available energy can be used to pump water to the upper reservoir (recharge phase), and energy is released when water moves back down to the lower reservoir through the turbine (discharge).

Pumped reservoir, Grand Coulee Dam, WA.

Image credit: Robert Ashworth via Flickr

Closed loop PSH storage does not need to be connected to an outside natural body of water, and all the water is re-circulated.

This storage technology is not new. The first commercial systems employed for storage were implemented in the 1970s, and the design changed very little since then. According to U.S. DOE, pumped-storage currently accounts for 95% of all utility-scale energy storage in the United States. However, additional investments are considered in innovative pumped storage technologies to explore its potential for storing non-dispatchable renewable power generated from utility scale wind and solar farms and improving grid resiliency and reliability.

Listen to the recent podcast that discusses the potential of various energy storage options for utility scale renewables: 

What to do you think? Which method of those mentioned in the discussion has a better chance to become the main player for storing mega- and giga-amount of power?

In this section, we will discuss how solar energy can be stored in the form of hydrogen gas. Hydrogen (H2) is a common industrially used chemical and fuel, which can be obtained from water by electrolysis or by reforming of natural gas. Electrolysis is of special interest in the energy storage context, since it converts electric energy into something storable. The process of electrolysis involves passing electric current through water or another aqueous solution, which initiates the electrochemical reaction:

$${\mathrm{H} }_{2 }\mathrm{O} \iff {\mathrm{H} }_{2 }+ 1 / 2 {\mathrm{O} }_{2 }$$

The basic idea is that the electricity generated by solar PV systems during daytime can be used to run electrolyzers to split water into hydrogen and oxygen gases. Hydrogen is collected and stored in one or another form. When energy is needed, hydrogen can be used for combustion or for electrochemical conversion (in a fuel cell) to recover energy as heat or electricity. Hydrogen provides a new form of energy economy, which complies with the present-day environmental requirements. For instance, hydrogen combustion does not result in any carbon emissions, and water and heat are the only products. Electrochemical utilization of hydrogen in fuel cells is thermodynamically efficient and environmentally benign. Fuel cells can be used for both stationary power generation and transportation. Unlike other forms of energy storage, hydrogen can be transported and used at a different location.

There are a few advantages of the hydrogen energy storage in solar plants:

1. Hydrogen generation by electrolysis is a well-established technology. Hydrogen is used in multiple branches of industry, so the procedures for its handling are well developed.
2. Water electrolysis uses low-voltage DC current, which is compatible with the output from the PV cells.
3. Hydrogen can be stored with minimal losses.
4. Hydrogen can have multiple uses - electricity generation, heat and power generation.
5. Hydrogen is an environmentally benign substance, its combustion does not produce harmful emissions, it is volatile and is easily dissipated.
6. There are multiple ways of hydrogen transportation (pipelines, tanks, hydride compounds).
7. Existing infrastructure for natural gas can be adapted to hydrogen use.

Electrolysis

There are different types of electrolysis that can be used for hydrogen generation. All of those methods use the available electricity to drive otherwise non-spontaneous electrochemical reaction. Electrochemical cell is required to realize electrolysis process. A typical electrochemical cell for electrolysis consists of two electrodes connected to an electric circuit submerged in the working solution. The electrode compartments are separated from each other with a gas-impermeable membrane, which does not allow hydrogen and oxygen mix, while allowing conduction of ions.

Use the following reading material to learn the scientific background and engineering principles for different electrolyzers.

Note on hydrogen safety

Due to high reactivity, hydrogen storage brings up some safety issues. At certain conditions, reaction between hydrogen and oxygen can lead to explosion (Figure 9.4), so storages for those gases should be separated, and leak detection is critical.

Figure 9.4. The character of the reaction of hydrogen with oxygen.

Credit: Shriver and Atkins, 1999

Solar hydrogen is attractive due to its carbon-free footprint, in contrast to any hydrocarbon fuels. Hydrogen has a very high energy density (142 MJ/kg), which is much higher than that of gasoline (~47 MJ/kg). This is due to its very light atomic mass. On the other side of the medal, hydrogen as a volatile gas must be contained in a compressed form, and the traditional gas cylinders used for storage and transportation of hydrogen have been a concern. Such cylinders, usually about 35 kg in weight contain only 700 g of hydrogen (at 200 bar of pressure), which is equivalent to only 2.85 liters of gasoline (Komoto et al., 2009).

The production of hydrogen via water electrolysis is most relevant to utility-scale PV systems. The efficiency of the electrolyzer is above 75% at optimized conditions. Even higher efficiencies are promised at elevated pressure, although in this case extra energy needs to be spent for isothermal compression. Much more than production, hydrogen "packaging", i.e., converting it to a convenient form for storage, can be much more costly and energy-expensive. These forms include (1) compressed hydrogen, (2) liquefied hydrogen, and (3) metal hydrides.

In option (1), compressed hydrogen is much more usable compared to that at atmospheric pressure due to much higher energy density (2.54 MJ/liter at 200 bar versus 0.0127 MJ/liter at 1 atm) (Komoto et al., 2009). Further compression to 800 bar brings its energy density to 10.1 MJ/liter - close to that of liquid hydrogen. However, multistage compression is estimated to take up to 8-12% of the total energy that hydrogen contains as fuel (referring to higher heating value), and mechanical and electrical losses can increase this value even further. Option (2), hydrogen liquefaction requires even more energy - 30 to 50 MJ/kg hydrogen converted. Option (3), metal hydrides are special compounds that can bind hydrogen by forming chemical bonds under certain temperature and pressure conditions and release hydrogen when temperature and pressure change. Simple physical pressurizing hydrogen into hydride form is possible, but is not very attractive due to the large amount of metal hydride needed (50 kg per 1 kg H₂ stored) (Bossel, 2006). Chemical binding of hydrogen in metal hydrides is much more promising.

Some examples of hydride compounds are CaH₂, MgH₂, LiH, NaAlH₄, LiAlH₄. Because metal-hydrogen bonds are quite strong, energy input is needed to release hydrogen from hydride. So, temperatures for hydrogen generation need to be raised to 120-200 ^oC. Nevertheless, the hydride storage is theoretically more efficient. For example, if we consider the reaction

$${\mathrm{CaH} }_{2 }+ 2 {\mathrm{H} }_{2 }\mathrm{O} => \mathrm{Ca} ( \mathrm{OH} {)}_{2 }+ 2 {\mathrm{H} }_{2 }- \text{ reaction of hydride with water, }$$

we can see that by molar ratios, 42 g of CaH₂ release 4 g of H₂ (gas), so it is roughly 4:1 ratio; at the same time, compressed storage in the cylinder provides storage of only 1 kg of hydrogen per 50 kg of storage tank, i.e., 50:1 ratio. Because of the obvious advantage of the metal hydride, substantial research effort is underway to improve this technology.

In this lesson, we looked at very different technologies used to store solar energy. Those technologies differ in physicochemical principles, scale, and impact, and require quite different scientific and engineering background for detailed analysis. In that sense, this material is challenging. At the same time, it is important for solar specialists to be aware of various energy storage options, as well as recent innovations, and to be able to apply those options to specific conditions. Hopefully, this lesson has been a good step in that direction and will motivate you to learn more and to elevate your expertise down the road.

This week, Lesson 10 continues the topic of energy storage – here we will try to identify the key issues related to the storage scale-up, when very large quantities of energy need to be stored for a relatively long period of time. In this lesson you will review the key metrics that help us compare different storage options and will apply those metrics in a storage sizing problem. This short, but useful exercise allows you to bring your understanding of battery parameters in a practical plane.

The discussion in this lesson is dedicated to the case study of PV+storage system operating in Hawaii – there are some good lessons to learn from the emergency situations on the Kauai Island demonstrating the ability of battery arrays to respond to blackouts. 

Purpose and Function of Utility Scale Storage Systems

Energy storage technologies are expected to enable of electric grid modernization, addressing the current limitations of electricity infrastructure and increasing grid stability and resiliency. We can identify a number of critical functions that we expect the energy storage systems to perform.

• Integration of renewable energy into the electricity grid

The variability of solar and wind power makes it hard for electricity providers to plug them into the electricity grid. Grids constantly balance the supply and demand of electricity and thus benefit most from dispatchable sources of energy (so far fossil fuels that could be burned on demand provided that sort of convenience). Energy storage makes the solar and wind energy more dispatchable (available on-demand of grid operators) and hence more competitive with traditional fuel options.

• Addressing peak demand

Responding to peak demands requires the ability to generate power quickly. The traditional choice for peak power generation are natural gas turbines. The energy stored in the batteries is immediately available and can be used to meet peak demand. This helps use the renewable power for peak generation and avoid grid disruptions or blackouts.

• Time shifting

PV Solar panels generate power only during the daytime, with the peak at noon hours, while the peak energy demand is often located during evening hours, when the solar irradiation is low. So the solar power needs to be “time-shifted” to be available during the time of high demand, and this can be achieved by means of utility scale storage.

• Energy autonomy and independence

For communities living in areas without access to electricity grid, combined renewable energy plus storage systems may be the best option to provide for constant supply of electricity. This autonomous approach can be realized at both distributed (house / community) and utility (area / region) scales.

Check out Table 10.1, which provides more detail on how storage systems serve the grid and help diversify the energy resources.

These purposes and applications require storage systems of diverse scale. The concept of scale has two dimensions: space and time. Space scale is related to the size and capacity of the storage, while timescale indicates how long the energy can be stored. Take a look at the figure below, which presents rough classification of storage systems in terms of size and time.

Figure 10.1. Typical time and size scales associated with energy storage technologies

(Aneke and Wang, 2016)

In this diagram, we see that such devices as capacitors store small amount of energy on the scale of seconds and minutes. At the same time, the systems shown in the right upper corner - pumped hydro storage, chemical storage - can store amounts of energy worth of gigawatt-hours over long periods of times (months to years). The same as with power generation technologies, storage system variety is important to satisfy various applications and demands and also to provide service storage in diverse natural and industrial settings.

The selection of the energy storage depends on many technical characteristics (besides scale), which would help us to understand why some technologies are preferred over others, and what trade-offs are involved in this selection. Let us look at some key storage characteristics next.

Key Metrics and Definitions for Energy Storage

There are a few key technical parameters that are used to characterize a specific storage technology or system. Those characteristics will determine compatibility of the storage with a proposed application and will also have impact on its economic feasibility. Let us go through some definitions.

Storage Capacity

Capacity essentially means how much energy maximum you can store in the system. For example, if a battery is fully charged, how many watt-hours are put in there? If the water reservoir in the pumped hydro storage system is filled to capacity, how many watt-hours can be generated by releasing that water? Those amounts are determined by storage capacity.

Understandably, the capacity of any storage will increase with the system size. The more battery stacks are installed, the more electric energy can be put in for storage. The larger the water reservoir, the greater energy turnaround becomes possible. The system size should be matched with the load and specific application.

Storage capacity is typically measured in units of energy: kilowatt-hours (kWh), megawatt-hours (MWh), or megajoules (MJ). You will typically see capacities specified for a particular facility with storage or as total installed capacities within an area or a country.

Sometimes you will see capacity of storage specified in units of power (watt and its multiples) and time (hours).

For example: 60 MW battery system with 4 hours of storage. What does it mean?

60 MW means that the system can generate electricity at the maximum power of 60 MW for 4 hours straight. That also means that the total amount of energy stored in the system is:

60 MW x 4 hours = 240 MWh

But it can also provide less power if needed. For example, if the load only requires 20 MW, the system can supply it for 12 hours. The total amount of stored energy is the same, but it is used more slowly:

20 MW x 12 hours = 240 MWh

So power and time ratings give us a little bit more information: we not only know how much energy is stored, but can also define at what maximum rate this energy can be potentially used.

Energy density

Energy density is often used to compare different energy storage technologies. This parameter relates the storage capacity to the size or the mass of the system, essentially showing how much energy (Wh) can be stored per unit cell, unit mass (kg), or unit volume (liter) of the material or device.

For example, energy densities for different types of batteries are listed in the table below [IES, 2011]:

Of course, we are interested to store as much energy as possible while using as small and light device as possible for this purpose. From the table above we can conclude, for example, that a fully charged Lead-Acid battery will run out of charge much sooner than a fully charged Li-ion battery of the same mass/size.

Energy density is related to capacity and determines the duration of power generation. Also materials with higher energy density help make the power block more compact, which is useful in portable electronics and vehicle applications.

Just for comparison, the energy density of the pumped hydro storage is 0.2—2 Wh/kg, which is rather low and requires significant masses of water and large reservoir size to deliver utility scale power.

Power density

Power density (measured in W/kg or W/liter) indicates how quickly a particular storage system can release power. Storage devices with higher power density can power bigger loads and appliances without going oversize. Imagine an electric vehicle accelerating from 0 to 60 MPH – which takes a lot of power. If you look at the table below, you will see why Li-ion battery remains the technology of choice for powering electric vehicles, even though some other battery types exhibit similar energy densities.

Figure 10.2 Classification of energy storage systems by energy and power density. Key to abbreviations is provided below.

Click for the key and a text description of Fig 10.2.

The image is a graph that displays the classification of energy storage systems based on energy and power density. The graph is a logarithmic scatter plot with 'Energy Density, Wh/liter' on the horizontal axis ranging from 1 to 10,000 Wh/liter, and 'Power Density, W/l' on the vertical axis ranging from 1 to 100,000 W/l. Different energy storage technologies are represented as colored rectangles and squares plotted on the graph. The technologies are abbreviated and color-coded as follows: SMES (Superconducting Magnetic Energy Storage) is a green rectangle placed high on the power density scale but low on energy density. DLC (Double Layer Capacitor) and FES (Flywheel Energy Storage) are placed at moderate levels of both energy and power density. Li-ion (Lithium-ion Battery), NiMH (Nickel Metal Hydride Battery), LA (Lead Acid Battery), NiCd (Nickel Cadmium Vented Battery), and NaNiCl (Sodium Nickel Chloride Battery) have varying positions but generally higher energy densities and moderate power densities. NaS (Sodium Sulfur Battery) and Me-Air (Metal Air Battery) exhibit high energy density but lower power density. PHS (Pumped Hydro Storage), CAES (Compressed Air Energy Storage), RFB (Redox Flow Battery), and HFB are on the lower end of both energy and power densities. H2 (Hydrogen storage) and SNG (Synthetic Natural Gas) have high energy density but low power density, with SNG depicted as a vertical bar on the far right of the graph. Two arrows are also included, one pointing to the right labeled "Decreasing volume" and another pointing up labeled "Increasing volume," indicating the relationship between volume and density in the context of the storage systems.

Credit: Mark Fedkin © Penn State based on the data from IES, 2011

The technologies located in the lower left corner of the diagram (low energy density and low power density) take significant amount of space and material to enable the storage conversion and are mostly suitable for very large scale projects. Systems such as PHS and CAES also rely on the availability of specific landscape and geological features to accommodate the storage reservoirs.

The technologies located in the upper right corner of the diagram are most coveted for portable and efficient power supply, such as electric vehicles. These compact systems can carry a significant amount of energy and release it quickly on demand.

The technologies in the upper left corner are special devices that can be used in quick response electronics. These systems store small amounts of energy (and therefore charging can be fast), but are able to provide high power by releasing energy within short period of time.

Finally, the technologies in the lower right corner are characterized by slow charge and discharge, but the advantage is the total high amount of energy they are able to store, providing longer duration of energy supply.

Storage efficiency

The main function of any storage device is to uptake and release power on demand. In case of a battery, for example, it would be electrochemical charge/discharge cycle; in case of pumped hydro storage, this process involves pumping water into the elevated reservoir and later releasing the flow through the turbine. Both charge and discharge processes include one or more energy conversions (Figure 10.3). In the figure, each arrow indicates the energy conversion from one form to another.

Figure 10.3 Main types of energy conversions in battery. 

Credit Mark Fedkin © Penn State

Figure 10.4 Pumped Hydro Storage systems

Credit Mark Fedkin © Penn State

Regardless the number of transformations, the energy comes to its initial electric form, which is finally ready to be dispatched into the grid. This is the charge-discharge cycle, the "round trip". 

In each conversion, energy is partially lost from the cycle and dissipated into the surroundings, and the efficiency of conversion at every step accounts for those losses. 

Efficiencies of all energy conversion steps in this cycle are combined in the metric called round-trip efficiency, which essentially indicates the percentage of energy delivered by the storage system compared to the energy initially supplied to the storage system. The obvious goal is to minimize the conversion losses and thus maximize the overall storage efficiency.

Here are some round-trip efficiencies of various energy storage systems:

These numbers mean the following. For example, out of 1 MWh of energy spent to pump water up to the hydro storage, only 0.7-0.8 MWh will be available to use after the water is released to run the turbine and generator to produce electric power. The other 0.2-0.3 MWh of energy will be converted into non-useful forms of energy and “lost” from the cycle. Some of the energy losses occur in the auxiliary devices used in the energy storage process, very often in the form of waste heat. Furthermore, energy losses may be linked to the mechanical or material losses: for example, leaks and evaporation of water from pumped storage, air leaks in CAES, chemical degradation and incomplete reactions in batteries.

The large-scale energy storage (also called grid energy storage) is a stand-alone or hybrid system that allows storing large amounts of electrical energy within an electrical power grid. Until recently, the dominant form of grid-scale energy storage has been pumped hydroelectricity, which accounts for over 95% of global installed storage capacity. Pumped hydro storage is rather old technology and has been around since the early 20th century, however it did not prove to be economically profitable or highly efficient. Other challenge you face in pumped hydro is the need of a certain kind of geological terrain to accommodate the proper size reservoir and availability of a large amount of water. We can add to that the environmental concerns and possible human impacts that are typically associated with the large hydro plants. It appears to be really hard to adopt this version of storage quickly in the newly constructed energy facilities.

Currently, there is a pressing need for new generation storage devices, that would be efficient, cheap, and possibly modular in order to facilitate their allocation at any location with any required capacity. This niche is currently being filled by the battery array systems. Battery arrays are often stand-alone facilities, strategically located to support regional grid stability.

PNM Prosperity Energy Storage Project, Albuquerque, NM.

Source: Sandia National Laboratory

To be clear, the large battery energy storage systems (BESS) are not huge batteries as a matter of fact. Battery arrays are modulirized systems, in which individual battery cells (for example, Li-ion batteries) are stacked in series into higher voltage units. The same as solar cells are combined in panels, and pannels are organized in arrays, scaling-up battery systems follows the same principles of series and parallel connections in order to achieve the required power and capacity.

Assembly principle of single battery cells into larger scale storage units. 

Source: AES Energy Storage

Some major companies that pursuing development and installation of large battery arrays are AES and Tesla. Follow the links below to learn about some case studies of battery array implementation:

• Tesla's Hornsdale Power Reserve in Australia:
  • Hornsdale Power Reserve
  • Tesla Beats Deadline, Switches on Gigantic Australian Battery Array
  • Tesla's Massive Powerpack Battery in Australia Cost $66 Million and Already Made up to ~$17 million
• AES Case Studies of Stationary Storage:
  • Case Study: AES Angamos BESS, Mejillones, Chile
• Angamos Storage Array
  • Energy Storage Experience in Chile & USA

One of the advantages we can see with battery solutions for grid storage is flexibility. For one, they can act as a reliable backup source. In the event of a power outage, battery systems can be turned on quickly to compensate. Data shown in the presentation on Chile grid system (second link for Angamos Storage Array) demonstrate very quick response of the arrays to the power outage and successful grid balancing.

In Australia, Hornsdale Power Reserve (built byTesla) helped restore power to the country’s grid in a fraction of a second after an unexpected failure at a power plant.

"the Hornsdale Power Reserve has smoothed out at least two major energy outages, responding even more quickly than the coal-fired backups that were supposed to provide emergency power.
Tesla's battery last week kicked in just 0.14 seconds after one of Australia's biggest plants, the Loy Yang facility in the neighboring state of Victoria, suffered a sudden, unexplained drop in output, according to the International Business Times. And the week before that, another failure at Loy Yang prompted the Hornsdale battery to respond in as little as four seconds — or less, according to some estimates — beating other plants to the punch. State officials have called the response time “a record,” according to local media." (Washington Post, 2017)

This table provides a list of other operating or commissioned projects, also including those based on other battery types (in addition to Li-ion):

Source: Zhang, Wei, Cao, Lin. Energy storage system: Current studies on batteries and power condition system. Renewable and Sustainable Energy Reviews. 2018 Feb; 82 (3): 3091-3106.

Batteries can also store extra energy. If there's excess power during peak wind or solar production, a battery can store up all that energy for future use. The Hornsdale plant is able to provide full power to 30,000 home, although for a relatively short period of time. In case of long outages, it still needs to be supported by traditional power generation facilities. In this way, battery essentially serves as an end user of power as well as a power plant.

According to National Renewable Energy Laboratory (NREL Report, 2018), the cost of the stand-alone utility scale Li-ion battery storage system breaks down as follows:

Figure 10.5. System costs of Li-ion battery utility scale energy storage for a stand-alone system.

Credit: NREL, 2018

Interesting to note that while the battery cost per unit energy remains the same (with battery system being modular), the costs related to the balance of system, installation, and service decline with the storage duration.

PV + Storage Systems

Co-locating the PV and storage systems has multiple benefits. Co-location results in cost savings by reducing costs related to site preparation, land acquisition, permitting, interconnection, installation labor, hardware (via sharing of hardware such as switchgears, transformers, and controls), overhead, and profit. The cost of the co-located, DC-coupled system is 8% lower than the cost of the system with PV and storage sited separately, and the cost of the co-located, AC-coupled system is 7% lower. (NREL Report, 2018)

Figure 10.6. Estimated costs of stand-alone and co-located PV and storage systems.

Credit: Energy.gov

DC-coupling or AC-coupling of the energy source with storage can be used in different scenarios. DC-coupling is less expensive due to fewer conversions - 1% lower total cost than AC-coupling, which is the net result of cost differences between DC-coupling and AC-coupling in the categories of solar inverter, structural balance of system (BOS), electrical BOS, labor, EPC (engineering, procurement, and construction) and developer overhead, sales tax, contingency, and profit.

“According to NREL, there’s only one utility-scale PV system in the United States connected to storage, and it is Lawa'i project in Kauai, Hawaii. There are more systems that have storage co-located with a solar array, but those batteries can be charged by other sources of power on the grid. According to GTM Research’s “U.S. Energy Storage Monitor 2017 Year in Review,” more than 5,500 energy storage systems are installed in the U.S., in the residential and commercial sectors with over 95% connected to PV in the residential sector at the end of 2017, which amounts to about 4,700 systems. By the end of 2018, GTM estimates that solar-plus-storage will have accounted for about 4% of distributed PV and could reach 27% by 2023.” (Energy.gov, 2019)

Based on this assessment, it looks like distributed PV market is somewhat ahead of the utility solar when it comes to hybrid PV+storage systems. It may change quickly. Look at the South Korea’s example:

“South Korea represents a story of how government planning can drive massive energy storage market growth, with a new policy to allow storage-backed wind and solar projects to earn renewable energy certificates worth five times their capacity value driving a massive boom in 2018. From less than 10 megawatt-hours deployed in 2017, South Korea’s utility-scale and commercial-industrial behind-the-meter deployments boomed to 1,100 megawatt-hours in 2018, with nearly $400 million in energy storage investments and a pipeline of projects that’s already overshot its goal of 800 megawatt-hours by 2020.” (JTM, 2019)

But let us take a closer look at the Lawa'i Solar and Energy Storage plant on Kauai (HI), and try to find out what conditions made commission of this project a success.

This article gives you some basic information on the setting and parameters of the plant. But let us not stop there. The project was commissioned in the beginning of 2019. I ask you to research some more information on its current status and share your findings on the class discussion thread.

Hawaii turned out to be a great "sandbox" for PV+storage projects after Lawa'i's success. It makes a lot of sense (both economic and environmental) with the state historically relying on petroleum shipments from the main land to generate its electricity, with the rates being some of the most expensive in the country (35-45 cent/kWh). More projects grew over the next five years, some summarized in the table below. 

Resource Constraints and Environmental Considerations

With energy storage industry changing fast, technology maturing, costs dropping, we can envision major shifts in both portable and stationary power markets. Cars and trucks will be powered by lithium-ion batteries rather than fossil fuel based internal combustion engines. Rechargeable lithium-ion batteries are also becoming crucial components of the new generation power grid to store increasing amounts of energy produced by solar and wind farms. This major technological scale-up will rely on the supply of new critical minerals and materials, such as lithium (Li), the lightest metal in the periodic table.

When thinking about this new future for our transportation and power industry, a few questions pop in our minds:

Where are we going to get all that lithium? (We are talking possibly 4-5 times topping the current market, perhaps more!) Is there enough of it on Earth? We are running out of oil, so would running out of lithium would get us in a similar trouble? How would prices for lithium affect international energy market and national economies of countries with higher or lower natural resource of this metal?

Companies such as Tesla, for example, already began a quest for control over lithium deposits across the globe, including countries in South America, Africa, and Australia.

Finally, mining and extraction of lithium at the accelerated rates will inevitably lead to serious environmental impacts on local and regional ecosystems (like any mining does), so how do these activities need to be regulated to keep this technology from becoming another global threat to biodiversity and human health?

To clarify some of these concerns, we are going to take a quick dive into the lithium lifecycle and learn about the key geological sources of its stock.

Mineral Spodumene is considered one of the key lithium ores and has been mined widely for multiple markets and applications.

Spodumene - lithium alumino-silicate, LiAlSi₂O₆. Mineral forms colorless, white, gray, yellow, light green, pink to purple crystals. Spodumene deposits are naturally associated with Li-rich pegmatites.

Credits: Rob Lavinsky, iRocks.com / S. Rae, Flickr

Prospecting for lithium and estimates of natural reserves have been done by U.S. Geological Survey at different years, and numbers keep changing due to discovering new deposits and, on the other hand, due to increasing demand for lithium from the energy storage and vehicle industry. For example in 2015, USGS estimated that world has lithium reserves for 365 years based on the global average production rate of 37,000 tons per year. Since then production has been doubled at least and the estimate shifted.

This is relatively recent (February 2019) USGS report, which provides some statistics on currently estimated reserves and production.

Countries that contain the largest lithium reserves include Argentina, Chile, Australia, and China. Among them, Australia is currently the leader in mining and production. You can make your own calculations based on the data provided and compare to some other online reports (but I am sure this information is not set in stone).

Next, answer a few self-check questions based on the above readings.

On the other side of the equation, we also need to understand the options for lithium disposal and recycling in the end of the battery lifetime.

Environmental Risks

Production and disposal of Li-ion batteries are associated with an array of environmental and health impacts, which include soil and water pollution due to open-pit Li mining, bio-toxicity, aquatic ecotoxicity (impact on fish), release of carcinogenic substances, high water and energy use. These impacts are not only associated with Li metal itself, but also with chemicals used to extract it from rocks or brines.

Figure 10.7. Li salt deposits at the Salar de Uyuni lake in Bolivia. Commercial extraction of Li from lithium brines takes a significant amount of water and energy.

Credit: Jan Beck via Flickr

Recycling of components and materials contained in Li-ion batteries becomes of paramount importance to mitigate two major issues: (1) resource depletion and (2) environmental pollution. Effective recycling programs would help reduce the need for opening new Li mines and offset the rate of lithium resource exhaustion at the national and global scales. Recycling routes would also help divert the battery products from landfills and thus avoid risks associated with chemical dispersion in soils, streams, and aquifers.

That said, the Li-ion battery recycling industry is really at its infancy still:

• Currently recycling of EV batteries consists of storage, landfill, and/or pyrometallurgical processes (burning in smelters).
• China - Legislated that all EV manufacturers and importers come up with a feasible recycling program.
• European Union - Has set a timeline for battery manufacturers and importers to recycle spent lithium-ion batteries.
• Canada - Has three provinces (British Columbia, Manitoba and Quebec) with mandatory recycling programs.
• USA - There are no Federal Regulations for battery recycling, some States do have.

(Source: Seeking Alfa, 2018)

For the next discussion assignment, we will try to search for signs of new business developments and regulations inside and outside of the U.S. related to Li-ion battery recycling and see how things have improved over the recent couple of years.

This lesson overviewed the main trends in the development of the large-scale energy storage systems. These systems are considered to be pivotal enabling technologies for greater penetration of renewables into the power grid. Cost, size, and ability to quickly install storage at any location on demand are key factors that will ensure the dispatchability of the new energy resources. Choosing the proper storage technology and sizing the system for a project can be a tricky balancing act between the ability to meet power requirements over a certain period of time and investment cost. In this lesson, you had a chance to study several successful operating solar+storage projects that provide us with a realistic picture of benefits as well as challenges of utility scale storage implementation.

Please go through the following activities to complete this lesson.