See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/282859641

ELECTRICITY FROM THE SUN

Conference Paper · January 2005

CITATIONS READS

2 220

4 authors, including:

Rudolf Bergermann Wolfgang Schiel

Schlaich Bergermann und Partner Schlaich Bergermann und Partner
23 PUBLICATIONS 538 CITATIONS 69 PUBLICATIONS 1,359 CITATIONS

SEE PROFILE SEE PROFILE

Gerhard Weinrebe
Schlaich Bergermann und Partner
58 PUBLICATIONS 1,176 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Stellio heliostat & heliostat fields View project

All content following this page was uploaded by Wolfgang Schiel on 15 October 2015.

The user has requested enhancement of the downloaded file.

 ELECTRICITY FROM THE SUN

Jörg Schlaich, Rudolf Bergermann, Wolfgang Schiel, Gerhard Weinrebe

Schlaich Bergermann und Partner, Consulting Engineers, Stuttgart, Germany
info@sbp.de

1. MOTIVATION

Current energy provision systems based on exhaustible fuels like coal, hydrocarbons and uranium are damaging
the environment and are non-sustainable. Increasing prices due to limited resources and rising demand pose
serious problems even to rich countries. Poverty, population explosion and migration are, amongst other reasons,
also a consequence of insufficient energy supply and high energy costs. According to a study of the international
Energy Agency about 16,000 million people have no access to electric energy, and about 2,4000 million have to
rely on biomass for cooking and heating with increasing tendency.
All these human beings want to and should be able to use electricity in the future. Hence demand for electricity
will dramatically increase, especially in developing and emerging nations. It would be very short-sighted to rely
on coal and oil or even nuclear power here. A clean inexhaustible source of energy is needed: The sun!
Which source could supply this energy without an environmental break-down (because most of these countries
have no funds for environmental protection) and without safety hazards (because most of these countries have no
safety history for nuclear power plants) and without rapidly exhausting their natural resources at the expense of
future generations?
The source is the sun! The less developed or poor countries frequently are the ones with high solar irradition and
vast deserts.
All this is widely known, everybody is talking about it, but much more should really be done.
Progress needs energy. (Fig. 1).
Corresp. to a doubling in years [a]

Yearly average growth of population [%] Energy consumption per capita [kWh/a]

14 5 120.000
Jord
Saud
CAN Norw
100.000
18 4 North America
Keni USA
Iran

Syri
Industrialized Countries 80.000
Ethi Swe
23 3 Nige Alge
Eastern Europe
Egyp
Vene
Ugan NL 60.000
Usbe Russia Trin Tsch
AUS D
Peru Isra
IND Mexi Bulg Western Europe
35 2 Ukra CH
GB F
Belorus Saud 40.000
C AUS J
Arge
Trin I I
Vene CAN
70 1 Alge
GR
E Isra
Usbe USA
Arge 20.000
Pol
Mexi GR NL CH
F J
C E ÄgypUkra Weiß E Norw
Cz Swe
SWE Iran Jord
GB
Aus Developing Countries D
0 0
0 5.000 10.000 15.000 20.000 25.000
low (65%) medium (25 %) high income ecconomies (10% of the world’s population of 5,5 Billion peoples)

Gross Domestic Product per Capita (US-$)

Fig. 1: Energy consumption and population increase in a country as a function of GDP or standard of living.
The higher a country’s standard of living, as measured by its per capita gross domestic product, the
higher its energy consumption and the lower its population growth. Thus, it would seem, if a poor
country had more energy economically available, its population would slow and its standard of living
rise!

October 2007 Page 1

ELECTRICITY FROM THE SUN SCHLAICH et all

In many regions of the world including India and China, two to four square kilometres of land is enough to
generate as much as 200-400 Gigawatt hours (GWh) of electricity per year using solar thermal technology. This
is equivalent to the annual production of a 50 MW conventional coal or gas-fired power plant. Worldwide, the
exploitation of less than 1% of the total solar thermal potential would be enough to stabilise the world climate
through massive CO2 reductions.
Thus solar electric generation becomes a new and promising field of activity – especially for structural engineers
as will be shown in the following.

2. TECHNOLOGIES FOR DIRECT SOLAR ELECTRICTIY GENERATION

When addressing solar electricity generation, it is common to distinguish between small decentralised power
generation and large centralised power plants. In the first case a unit has, say, an output of some 100 kW
maximum and is therefore likely to be installed in remote areas away from the grid. Centralised generation
means large-scale solar power plants feeding the grid with 10 to 500 MW per unit and capable of someday, at
least partially replacing present large-scale fossil or nuclear electricity generation. Bearing in mind the successful
operation of hydroelectric power plants, this is nothing less than an utopian goal.
According to the present state of the art, the following technologies for direct, active solar electricity
generation are widely known and generally have been found to be feasible:

Decentralised small-scale direct solar electricity generation

- Photovoltaics (PV)
- Dish/Stiring systems (D/S)

Centralised large-scale direct solar electricity generation

- Distributed collector systems (DCS) with parabolic trough collectors

- Central receiver systems (CRS) with heliostats
- Solar Updraft Towers

Solar Thermal Power Plants

Point focusing Line focusing Non focusing

Dish/Stirling- Central Receiver Parabolic Trough SolarUpdraft

Solar ChimneyTowers
Systems Systems Plants

Fig. 2: Solar Thermal Power Plants

With the exception of PV, all theses systems are solar thermal power plants (Fig. 2) (only wind power is
indirect). Both solar updraft towers and photovoltaics, as against those using reflectors, are also able to utilise

October 2007 Page 2

ELECTRICITY FROM THE SUN SCHLAICH et all

diffuse solar radiation. This suggests that the Solar Updraft Tower provides the only technology for large
scale utilisation in the regions with substantial diffuse global radiation, i.e. major parts of Asia including
India, China, South of Africa and Southern America.

3. PHOTOVOLTAICS (PV)

Solar cells are photoelectric or photovoltaics (PV) cells which at normal temperatures transform solar light into
electricity without any mechanical or moveable parts. PV cells are also very durable and reliable and can be
added in modules of any size and output, an ideal method at a first glance. However, in spite of immense
research efforts and considerable progress, the cost of PV electricity is still by far higher than that of any other
solar systems including wind technology. It must be expected that his will remain so for a long time.
For very decentralised mini-scale energy production and consumption – for SOS telephones, on boats or even
planes – PV is unique and already has achieved wide application.
In any case, PV does not present much of a challenge for structural engineering since the frames supporting PV
cells from a structural point of view are rather simple. Therefore PV will not be further discussed here.

4. DISH/STIRLING SMALL SOLAR POWER PLANTS (D/S)

Principle
The main components of a Dish/Stirling System are a parabolic concentrator, a tracking system, a solar heat
exchanger (receiver) and a Stirling motor with generator (Fig. 3).

Fig. 3: Schematic representation of a Dish/Stirling System

The concentrator tracks the sun bi-axially in a way that the optical axis of the concentrator always points to the
sun. The solar radiation is focused by the parabolic concentrator onto the solar receiver which is situated close to
the focal point of the parabola. The receiver feeds the high temperature thermal energy into the cycle of the
Stirling motor, which converts it highly efficient into mechanical energy. The directly coupled generator finally
converts it into electric energy. Dish/Stirling Systems are characterized among other solar thermal power plants
on one hand by their high efficiency and on the other hand by a high modularity in their application.

Stretched Metal Membrane Dish-Concentrator (Fig. 4)

Tension structures, especially if double curved, stand for an ideal load bearing behaviour. They may be
prestressed mechanically from their edges resulting in negative curvature or saddle shapes, or pneumatically
resulting in positive curvature or spherical shapes.
If, for better durability, they are manufactured from thin metal membranes (say 0.2 – 0.5 mm thick) instead of
the usual textile material, the common approach of shop fabrication by use of a predetermined cutting pattern
fails, since double curved metal sheet would crackle and spoil if folded for transport from the shop to the site. On

October 2007 Page 3

ELECTRICITY FROM THE SUN SCHLAICH et all

the other side, especially stainless steel sheet is extremely ductile and therefore the fabrication of double curved
surfaces from plane sheet through plastic deformation appears to be the right approach in this case. So, for
example a dodecahedron becomes a sphere through inflation (Fig. 5) as well as two plane sheets, welded from
strips and clamped between two compression rings, become a cushion (Fig. 6). Such inflated cushion could serve
as a light weight roof for a circular silo or an arena, if a small permanent internal pressure is maintained.

Fig. 4: Stretched Metal Membrane Dish-Concentrator

Whilst doing such tests with this purpose in mind several years back, we observed with increasing internal
pressure respectively strain of the sheet all welding wrinkles disappearing and a perfectly smooth spherical shape
evolving – a surface with optical quality! (Fig. 7) This lead to the idea of reversing the process, i.e. to evacuate a
short cylindrical drum thus producing a concave surface respectively two on either side one of which should be
fabricated with special care and could be used as a solar concentrator (Fig. 4 and Fig. 8). In fact for heliostats
(for CRS-systems, see later in this paper) with a long focal length the membrane remains in its elastic range. For
D/S-systems, however, the membrane is plastically deformed to the desired shape by air pressure and additional
hydraulic load (Fig. 8 and Fig. 9). Thin mirror glass is bonded to the reflecting surface. When the concentrator is
in operation, the shape of the membrane is kept constant by a partial vacuum in the interior of the concentrator.

Fig. 5: A dodecahedron welded from 12 equal plane stainless steel sheet pentagons (a) before inflation (b)
after inflation

October 2007 Page 4

ELECTRICITY FROM THE SUN SCHLAICH et all

Fig. 6: Metal membrane cushion formed from plane sheet by inflation

a) two sheets are clamped between two steel rings
b) the interior is inflated and the two sheets deform plastically

Fig. 7: The contour lines of the cushion at 500 mm rise of the sphere

water basin
(only for forming)
hydraulic
stretching ring
water

forming

Pi < Pa
concentrator
ring

front membrane back membrane

(after deformation)

Fig. 8: Forming the membranes of a dish concentrator by deflation and water loading
October 2007 Page 5
ELECTRICITY FROM THE SUN SCHLAICH et all

Fig. 9: Forming a stretched metal membrane concentrator by underpressure and hydraulic load
a) placing the second membrane
b) stretching the membrane
c) water loading and deflation
d) stretched metal membrane D/S-System: 9 KWel Systems in Almeria, Spain (SBP)

Segmented Sandwich Dish-Concentrator (Fig. 10)

1 Concentrator shell
2 Stirling support structure
3 Stirling unit
4 Elevation drive arc
5 Azimuth drive arc
6 Foundation
7 Azimuth drive
8 Switch case
9 Turntable
10 Concentrator ring truss
11 Elevation bearing

Fig. 10: Segmented Sandwich Dish-Concentrator

October 2007 Page 6

ELECTRICITY FROM THE SUN SCHLAICH et all

Detailed studies of the fabrication and costs of D/S-systems with metal membrane concentrators revealed that
they are the right solutions if the site installations for the fabrication can be made use of frequently, i.e. if series
of say 50 D/S units are to be installed in a cluster.

Fig. 11: Assembling a Segmented Concentrator

Therefore for single units a dish was developed which is composed on site of individual segments of a size
which can be brought on site in a commonly used container without any need for site installations. The
individual segments of the so-called EuroDish are sandwiches made from glass-fibre reinforced epoxy-resin.
After being brought to their location these elements are joined to become a spherical shell of considerable
strength and stability. Its reflecting concave surface is covered with 0.9 mm thin 30x50 cm glass mirrors as with
the stretched membrane, guaranteeing a reflectivity of about 94%
The optical quality and precision of the segmented concentrator does not at all fall behind that of the metal
membrane alternative.

The Energy Conversion System – Hybrid Mode

The energy conversion system consists of a Stirling engine with a receiver located at the focal point of the
concentrator. The reflected solar rays heat the working gas (hydrogen) of the engine. A generator is coupled
directly to the engine to produce electricity (Fig. 12).
Since each unit is capable of fully independent operation, as many concentrators as desired can be operated in
conjunction, according to requirements. They can be operated both in “grid connection” or “stand-alone” modes.

Fig. 12: Energy Conversion:

a) The principle of the Stirling engine
b) The 10 kWel power pack (SOLO): Receiver, Stirling and Generator

October 2007 Page 7

ELECTRICITY FROM THE SUN SCHLAICH et all

If power output is required independent from the existing meteorological conditions, in the evening or by night
the Dish/Stirling system can (besides the use of batteries) be configured as hybrid system. Hybrid means that
additional fossil energy sources (e.g.: bio gas) can be used to add thermal energy to the solar thermal energy to
stabilize the power output or for prolongation of operation time in the evening or at night. Hybrid Systems have
already been successfully tested in long-time operation.

System Properties and Application

Dish/Stirling Systems have been developed so far by the author and his team with power outputs ranging from
several kWel to 50 kWel and concentrator diameters up to 17 m. Recent developments are concentrated on
power output from 10 to 25 kWel. In this range both stand alone and grid connected systems were designed and
tested.
Due to the low thermal inertia of the energy conversion, a Dish/Stirling System reacts very quick on changes in
solar thermal input. Thus steady state operation is achieved already shortly after the start up of the system.
(Fig. 13)
The field of possible applications covers on the one hand the support of smaller or bigger grid systems. On the
other hand stand alone systems can power for example water pumps or desalination plants. If Dish/Stirling
systems are installed gathered in clusters, applications up to 10 MW can be realized. Above this limit other solar
thermal systems are more economical (see below).

20 10
net efficiency
18 Design 9

16 Power (gross) Efficiency 8

Power (net)
14 7
Efficiency [ % ]

Power [ kW ]

12 6
Power
10 5

8 4

6 3

4 2

2 1

0 0
0 100 200 300 400 500 600 700 800 900 1000
Direct insolation [W/m²]

Fig. 13: Input/output characteristics and net efficiency

0,70
Levelised Energy Cost [Euro / kWh]

At sites with high solar insolation Dish/Stirling

PV - 65 T€/kW
0,60
PV - 55 T€/kW
Diesel 0,40 €/l
0,50
Diesel 0,60 €/l

0,40 Diesel 0,80 €/l

Fig. 14: Calculated levelised energy
Diesel 1,00 €/l
costs for D/S-system as a
0,30
function of number of units
0,20 produced in comparison with
fossil-fuelled plants (Diesel)
0,10

0,00
1 10 100 1000
Interest rate: 6% PV Diesel
Depreciation time
time: Number of units per year
PV: 20 years
Diesel: 10 years
D/S: 10 years

October 2007 Page 8

ELECTRICITY FROM THE SUN SCHLAICH et all

As the fabrication of the concentrator is relatively inexpensive, such systems make economical electricity
generation a real possibility. (Fehler! Verweisquelle konnte nicht gefunden werden.)
With carefully planned technology transfer such power plants could readily be fabricated in China. On the other
hand, the design has to consider the effects of extreme thermal expansion and special high-temperature-resistant
materials have to be used for the receiver tubes in order to achieve long service life.

Recent developments and projects

All systems so far were built in single piece production and have therefore a high investment cost level.
Additionally Stirling motors still require regular maintenance. Important goals for further development are an
increase of reliability and further cost reduction.
The author and his team are working currently with several German partners supported by the German Ministry
of the Environment (BMU) on measures for cost reduction and on first steps into the market. As part of the
project demonstration plants were erected in Seville (Spain), Odeillo (France) and Würzburg (Germany)
(Fig. 15).

Fig. 15: 10 kW Country Reference units in (a) Spain, (b) Italy, (c) Germany and (d) France (SBP)

The next step will be a plant with 0.5 to 1.0 MWel respectively 50 to 100 Dish/Stirling systems (Fig. 16). With
this step the first series production can be achieved and at the same time the operational experience can be
extended. Comparable projects are also planned by competing groups in USA. A 1.0 MWel plant is planned to
be installed in Nevada. With series production, levelised energy cost between 0.15 and 0.2 €/kWh depending on
site conditions are expected. American studies predict for big series even 8 USCent/kWh for this technology. It
is expected that first markets can now be penetrated.

October 2007 Page 9

ELECTRICITY FROM THE SUN SCHLAICH et all

Fig. 16: Planned 0.5 MW dish Stirling plant (SBP)

5. DISTRIBUTED COLLECTORS SYSTEMS (PARABOLIC TROUGH)

Principle

The reflective surface of a parabolic mirror concentrates solar radiation onto a vacuum insulated receiver tube
which is located in the focal line (Fig. 17). The heat transfer fluid (HTF) in the tube is a thermo oil which is thus
heated up to 400°C. Mirrors and receiver tube are tracking the sun along one axis.
By serial connection of single collectors, loops with the length of several 100 meters can be generated. The hot
oil of all loops is collected and fed into the steam generator in the central power block. A conventional steam
turbine (steam temperature 370°, 100 bar) with generator converts thermal to mechanical and electrical energy,
respectively. (Fig. 18)
glass envelope

direct solar radiation

absorber tube
with selective
coating

Parabolic trough
concentrator-
with reflecting
surface
tracking

Fig. 17: Principle of a parabolic trough collector

October 2007 Page 10

ELECTRICITY FROM THE SUN SCHLAICH et all

Fig. 18: Rankine Cycle Power Plant layout for Electricity Generation in a hybrid-operated steam power-cycle
(below)

The structures supporting the curved mirror segments should be light, stiff and easy to manufacture. Future
aperture widths of 9 m represent challenging structural design problems.(Fig. 21)
In case of prolonged lack of solar radiation, a gas-burning capability may be added. This back-up avoids the
need for costly thermal storage and guarantees continuous electricity supply, requirement utilities certainly
appreciate and pay for correspondingly, though they disrespectfully call DCS just fuel-savers.

Fig. 19: Structural model of single trough collector element (68 m² collector surface)

October 2007 Page 11

ELECTRICITY FROM THE SUN SCHLAICH et all

Fig. 20: Structural Details of EuroTrough

Installed Systems
In the Majave desert of California nine Parabolic Trough Power Plants produce electricity from the sun since
almost 20 years. All SEGS (Solar Energy Generating System) Power Plants have a nominal capacity of 354
MWel (Fig. 21). Fossil heating (natural gas) guarantees a constant output. The first plant (SEGS I) with 14
MWel was put in operation in 1984, the last plant (SEGS IX) with 80 MWel started operation in 1990. In this
development many details were optimised and efficiency was increased. In parallel operation and maintenance
cost were decreased significantly.

Fig. 21: Arial view of the 250 MW power plants SEGS IV – VII at Kramer Junction, USA

October 2007 Page 12

ELECTRICITY FROM THE SUN SCHLAICH et all

Recent Developments and Projects

After almost ten years of standstill in the Parabolic Trough technology the development was intensified again in
the 90s. Several prototypes and demonstration units were erected to prepare the realization of further commercial
plants.
A European consortium developed a weight and cost optimised collector, the author’s team of Schlaich
Bergermann und Partner was responsible for design and engineering within the consortium.
In the years 2000 and 2001 two prototypes (EuroTrough I and II,) were erected and tested on the Plataforma
Solar de Almeria in Spain with financial support of the European Commission. (Fig. 22)
Main characteristic parameters of EuroTrough 100 m and 150 m
EuroTrough Model ET 100 ET 150

Focal Length 1.71 m 1.71 m

Absorber Radius 35 mm 35 mm

Aperture Width 5.76 m 5.76 m

Aperture Area 552 m² 828 m²

Collector Length 99.5 m 148.5 m

Number of 12 m Modules per Drive 8 12

Number of Glass Facets 224 336

Number of Absorber Tubes 24 36

Mirror Reflectivity 94% 94%

Weight of steel structure and pylons, per

19.0 kg 18.5 kg
m² aperture area

Fig. 22: Prototype of a EuroTrough (72 m in length) at Almeria, Spain

October 2007 Page 13

ELECTRICITY FROM THE SUN SCHLAICH et all

Encouraged by the results the partners Schlaich Bergermann und Partner (SBP), Solar Millennium AG and
FlagSol started to develop a 50 MWel commercial plant in Spain. To reduce the up scaling risk from the
prototype (410 m²) to the power plant (500,000 m²), a demonstration loop with 4,360 m² surface was planned as
an intermediate step.
This demonstration loop is in continuous operation as an integrated component of SEGS V at Kramer Junction,
USA since 2003.(Fig. 23)

Fig. 23: Demonstration Power loop (2x400 m loop) at Kramer Junction in California, USA (SBP et al.)

The economic and technical analysis of the demonstration unit certifies the collector to be a real improvement
compared to the existing designs. The material quantities could be reduced by increasing the stiffness of the
structure. Optimised alignment procedures have led to a higher collector efficiency. The on site erection was
improved. A new receiver tube developed by Schott, Germany, with a highly selective coating was implemented
in 2004 to further increase the efficiency.
Besides the classic Parabolic Trough systems there are alternative developments under way. On Fresnel
Collectors the parabolic surface is divided into individually tracked stripes of mirrors installed close to the
ground. Thus the Receiver tube can remain at a fixed position but requires a secondary concentrator to achieve
the necessary efficiency. A demonstration unit in a technical scale is the next step for this technology.

Recent Projects

The Solar Millennium AG, Schlaich Bergermann und Partner and FlagSol together with ACS Cobra, Spain, are
at present installing three power plants close to Guadix, Spain.. A thermal storage will allow 6 h of operation at
nominal power output (night operation). The parabolic trough plant is supposed to produce app. 50 MWel and
will dispose of a thermal salt storage system. The total collector aperture area is 510,000 m², the turbine inlet
conditions 100 bar/370°C. Each plant will feed approximately 160 GWh per year of pure solar electricity into the
Spanish grid.

6. CENTRAL RECEIVER SYSTEMS (CRS) WITH HELIOSTATS

Principle

Characteristic for CRS plants is a field of reflecting mirrors (heliostats) that focus solar radiation onto a central
receiver that is positioned on a tower in the middle of the heliostats. (Fig. 24, Fig. 25) The heliostats are
parabolic or spherical shaped mirrors that track the sun bi-axially. At the receiver, the solar energy is converted

October 2007 Page 14

ELECTRICITY FROM THE SUN SCHLAICH et all

into thermal energy and a heat transfer fluid (air, molten salt, water/steam) feeds it into a conventional steam
turbine cycle that finally drives a generator (Fig. 26). To get constant steam parameters also during cold passages
either a heat storage system or additional heating (e.g.: with natural gas) with fossil fuels is foreseen.
Prototype testing plants up to 10 MWel were installed in France, Israel, Italy, the former USSR, Spain and the
USA. Commercial CRS plants are not in operation up to now.

Fig. 24: The 10 MW Central Receiver system Solar Two near Barstow, CA, USA

Receiver

Concentrated solar radiation

Direct solar insolation Direct solar insolation

Heliostat field Tower Heliostat field

Fig. 25: Principle of a CRS plant

October 2007 Page 15

ELECTRICITY FROM THE SUN SCHLAICH et all

Air-Receiver
Burner
Concentrated 700°C
Radiation Steam
Generator
Turbine

G
Heliostat Field
Air Cycle Steam
Cycle
Pump
200°C

Condenser
Blower

Fig. 26: Scheme of a CRS Plant with air as a heat transfer fluid

Heliostat Field

The Heliostat field consists of up to thousand individual heliostats. The concentrator surfaces of heliostats range
between 40 and 150 m². A maximum of 200 m² was installed. The heliostat field causes more than 50% of the
investment costs of the plant. Therefore considerable efforts were undertaken to develop economical heliostats
with high optical efficiency and high reliability. Two different design principles can be identified: Glass/metal
facetted heliostats and Glass/metal membrane heliostats.(Fig. 27)

Fig. 27: Facetted Heliostat (left) and metal Membrane Heliostat (right)

Facetted Heliostats

Their concentrator consists of a steel lattice structure installed on a torque tube, supporting a multitude of
rectangular single mirrors, the so called facets, with a size of 2 to 3 m² each. The tracking unit comprises a
vertical pedestal bolted to the foundations with an azimuth/elevation gearbox on top, to which the torque tube is
connected (Fig. 28).

Membrane Heliostats

To overcome the high construction and installation costs associated with facetted designs, and to achieve a
higher optical quality at the same time, heliostats on the basis of stretched metal membranes were conceived by
the author and his team of Schlaich Bergermann und Partner. Their concentrator corresponds with that of
stretched Metal Dish made of drums with thin metal membranes stretched like drumheads (Fig. 30). Thin glass
mirrors are glued to the front side stretched metal membranes to achieve the desired reflectivity. Using a small
fan or pump, a slight under-pressure is maintained in the concentrator plenum. Thus the membrane deforms
elastically, and the flat mirror becomes a concentrator. Using justifiable effort, the optical quality achieved with
large stretched membrane heliostats is significantly higher than that obtained with glass-metal facetted heliostats
October 2007 Page 16
ELECTRICITY FROM THE SUN SCHLAICH et all

of comparable size. An additional advantage is the fact that by varying the under-pressure inside the
concentrator, the focal length of the heliostat can easily be adjusted during operation. Thus it is unnecessary to
manufacture, install and adjust special facets for heliostats for each focal length respectively distances between
heliostat and receiver, as it is the case with facetted heliostats. One concentrator satisfies all heliostat positions.
Schlaich Bergermann und Partner have designed and built stretched metal membrane heliostats with 44 m² and
150 m² (Fig. 28, Fig. 30). Both are characterized by an excellent optical quality. For tracking of the membrane
heliostat, the azimuthal tracking as for the Dish/Stirling is used again. This results in much more stability under
wind as with the facetted heliostats and their vertical podestal. So membrane heliostats guarantee many more
operation hours and electricity production.

Fig. 28: A conventional faceted heliostat Fig. 29: A first stretched metal
membrane heliostat, 45 m²

Fig. 30: The 150m² Advanced Stressed Membrane heliostat ASM150 at the PSA, Spain.

Other components

Other components like steam circuit, balance of plant, control and instrumentation are very similar to those used
in fossil-fuelled power plants.

Current projects

With the Spanish feed-in tariff for solar electricity of roughly 0.22 €/kWh as an incentive, two CRS projects are
being developed:

October 2007 Page 17

ELECTRICITY FROM THE SUN SCHLAICH et all

PS10. Based on the positive experience gained at the Spanish 1 MW CRS Test Facility at the Plataforma Solar
de Almeria a 11 MW CRS plant has been constructed in southern Spain near Seville in 2006/7. A saturated
steam receiver on top of a 100m high tower with 4 panels 5.4 m x 12 m will be used. The heliostat field consists
of 624 heliostats with 121 m² each. The steam process is designed for 250°C and 40 bars and a yearly energy
production of 23 GWh is expected.
Solar Tres. This project is based on the know-how and experience gathered during design, construction and
operation of the 10 MW Solar Two system near Barstow, California. Hence the name “Solar Tres” (Spanish for
“Solar Three”). The 10 MW Power Tower design comprises a molten salt tube receiver and an integrated
thermal salt storage, enabling continuous operation of the plant.

7. SOLAR UPDRAFT TOWERS

Principle

The solar updraft tower combines three well-known technologies – the greenhouse, the updraft tube and the
turbine – in a novel way (Fig. 31, Fig. 32). Incident solar radiation heats the air under a large transparent
collector roof. The temperature difference causes a pressure drop over the height of the tube resulting in an
upwind which is converted into mechanical energy by turbines and then into electricity via conventional
generators.

Fig. 31: Drawing of several large (100 – 200 MW) solar updraft towers in a desert

October 2007 Page 18

ELECTRICITY FROM THE SUN SCHLAICH et all

Solar Radiation

Tower

Turbines
Solar Collector

Fig. 32: Principle of the Solar Updraft Tower: Glass roof collector, tower and turbines

This solar energy system has many technological and physical advantages:
- Global radiation, including diffuse radiation when the sky is overcast, can be exploited.
- The natural storage medium – the ground – guarantees operation at a constant rate until well into the
hours of darkness (and throughout the night with large-scale installations). If in addition black water-
filled tubes are placed on the ground underneath the roof, a continuous 24 hours electricity production
can be achieved (Fig. 33)
- There are no moving parts, nor are there parts that require intensive maintenance aside from the turbine
and the generator.
- Not even water is required to cool mechanical parts.
- its simple, low-cost design and materials (glass, concrete, steel) make solar updraft tower systems
applicable to any desert country, since glass and concrete are made of sand and stone which are
available anywhere. Labour represents a high portion of the installation costs. This will stimulate the
local labour market, while at the same time helping to keep overall costs down.
100
natural ground storage
water storage 10 cm
80
water storage 20 cm
Power (%)

60

40

glass roof
20
into the soil and
into the air into the air
the water tubes

0
0:00 3:00 6:00 9:00 12:00 15:00 18:00 21:00 0:00
soil soil
water tubes
Day Night Time (h)

Fig. 33: Left: principle of heat storage with water tubes under the collector roof,
Right: daily net power output depending on the thickness of the water layer

There is in fact no optimum physical size for solar updraft towers. The same output may result from a large tube
with a small collector roof area and vice versa. Thus, to decide the optimum dimensions of height of tube against

October 2007 Page 19

ELECTRICITY FROM THE SUN SCHLAICH et all

collector radius, the specific construction costs of these items must be known. If glass is cheap but concrete
expensive, a large collector and low tube is preferable, and vice versa (Fig. 34). Broadly, to achieve a maximum
output of (30) 200 MW at an irradiance of 1,000 W/m², the roof must have a diameter of (2,200) 4,000 m, if the
tube has a height of (750) 1,500 m. If black water-tubes are placed on the soil underneath the roof for continuous
200 MW full load 24 hours electricity production the diameter of the roof must be increased to 7,200 m. Now
this solar updraft tower from a solar radiation of 2,300 kWh/m²a extracts about 1,500 GWh/a, in fact a power
plant!

Pel = Gh * AKollektor *ηKollektor *ηKamin *ηTurbine

g
ηKamin = H Kamin
c p ∗ T0
Pel ~ AKollektor * H Kamin

Fig. 34: Yearly net energy production of Solar Updraft Tower Plants versus collector diameter and height of
tube

The collector roof, responsible for more than 50% of the total cost must be as economical as possible. For that
the glass panels are placed on suspended stress ribbons made from steel slats, spaced 1 m. They are supported by
underslung girders resting on steel tubular columns 9/9 m². Tests on a prototype solar updraft tower in
Manzanares / Spain has shown that this is a most efficient and durable structure (Fig. 35, Fig. 36).

Fig. 35: The solar updraft tower prototype at Manzanares / Spain

October 2007 Page 20

ELECTRICITY FROM THE SUN SCHLAICH et all

Fig. 36: The glass collector roof of the Manzanares solar updraft tower

The turbines are basically more closely related to the pressure-induced water turbines than to the velocity-
induced natural wind power plants. Therefore, the builders of hydroelectric power plants aided us in the
development and cost estimate. Either several horizontal axis engines are placed around the base of the tube
(Fig. 32) or – the cheaper solution – one large, say 200 MW turbine with a vertical axis is placed in the tube’s
diameter (Fig. 37).

~20 to 40m

Fig. 37: 200 MW vertical axis turbine in the shaft of the solar updraft tower

For the tube itself we thoroughly compared the possible construction methods and the materials such as covered
steel framework with cable nets, membranes, trapezoidal metal sheet etc., just to discover that for all countries in
question the reinforced concrete tube promised the longest life-span at the most favourable costs (Fig. 38).
Technologically speaking they are nothing but cylindrical natural draught cooling towers with – as shown in
Fig. 39 as an example – a diameter of 120 m and height of 1,000 m. The wall thickness decreases from 99 cm
just above the support on radial walls to 25 cm halfway up, then remaining constant all the way to the top. Such
thin-walled tubes will oval due to the wind suction especially at the flanks (Fig. 40). This tremendously
increases the meridional compressive and tensile stresses if compared with the linear bending stresses of a
cantilevering beam (Fig. 41, top left). The resulting loss in stiffness due to cracking of the reinforced concrete
and the danger of buckling limit the height of natural draught cooling towers to about 200 m. But this ovaling
can be efficiently counteracted by stiffening spoked wheels which have the same effect as diaphragms, hardly
affecting the upwind. If the spokes are made of vertical steel slats stressed between the compression ring along
the tube’s wall and a hub ring, such a spoked wheel is prestressed by its own weight, thus resulting in tension
and compression resistant spokes (Fig. 42). It is seen from Fig. 41 that the meridional stress in the chimney wall,
shown in the diagrams across the diameter and the height, do undulate tremendously without any spoked wheels.

October 2007 Page 21

ELECTRICITY FROM THE SUN SCHLAICH et all

But one spoked wheel at the top and another one or even three more further below do reduce the meridional
stresses to an extent that tension disappears completely, succumb by the tube’s dead load. Considering that the
absolute volume under these stress diagrams is somehow proportional to the consumption of concrete and
reinforced steel, one arrives at the obvious conclusion that these spoked wheels make such high towers for solar
tubes feasible.

120

1000 0.30m
170
outside

intside

180

550 0.30m
180

500 0.31m
180

300 0.47m
225

55 1.80m
30 0.80m
55

Fig. 38: left: Construction of the updraft tube with jump form

October 2007 Page 22

ELECTRICITY FROM THE SUN SCHLAICH et all

Fig. 39: right: Wall thickness of a tube 1,000 m high and 120 m in diameter

Cp = + 1.00

Cp = - 0.45
wind direction

pressure
Cp = - 1.5
suction

Fig. 40: Typical pressure distribution around the circumference of a cylindrical tube

without spoked wheels spoked wheel at 1000 m

Meridian stress Meridian stress

[MPa] [MPa]

45 45
30 30
15 15
0 0
-15 -15
-30 -30
1.000

1.000
900
900
800

800

-45 -45
700
700
600

600
500

500

-60 -60
400

400
300

300
200

200
85

65

15
-15
80

-43
-65
-80
100

85
80
65

15

-65
-80
100

height [m]
43

-15
-43

height [m]
43

diameter [m] diameter [m]

spoked wheels at 500 and 1000 m spoked wheels at 500, 600, 850 and 1000m

Meridian stress Meridian stress

[MPa] [MPa]

45 45
30 30
15 15
0 0
-15 -15
-30 -30
1.000

1.000
900

900
800

800

-45 -45
700

700
600
600
500

500

-60 -60
400

400
300

300
200
200
-15

15
-15
85
80
65

15

-43
-65

100

85

65

-65

100

height [m] height [m]

43

-80

80

-43

-80
43

diameter [m] diameter [m]

Fig. 41: Meridional stresses in the updraft tube according to Fig. 39 around its periphery and along its height
depending on the number of stiffening spoked wheels

October 2007 Page 23

ELECTRICITY FROM THE SUN SCHLAICH et all

Fig. 42: Spoked wheels, the spokes are made of vertical steel slats

0.50 €

0.45 €
Levelized Electricity Cost in €/kWh

0.40 € 5 MW

0.35 €

0.30 €

0.25 €

0.20 € 30 MW

0.15 € 100 MW
200 MW
0.10 €

0.05 €

0.00 €
1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12%
interest rate

Fig. 43: Electricity cost as a function of size of Solar Power tower (MW) and interest rate

October 2007 Page 24

ELECTRICITY FROM THE SUN SCHLAICH et all

Fig. 44: Electricity costs from solar updraft tower, coal and combined cycle power plants according to the
present business managerial calculation (overall interest rate: 10.7%)

Thus, with the support of construction companies, turbines manufacturers and the glass industry a rather exact
cost estimate for a 200 MW solar updraft tower could be compiled. (Fig. 43, Fig. 44)
In case of the solar updraft tower the interest on the investment governs the price of electricity, whereas in the
case of fossil fuel power plants mainly the fuel costs are the deciding factor.
With increasing fuel costs the solar updraft tower becomes more and more favourable.
The presently still higher costs of solar electricity are balanced by several advantages:
- no ecological damage and no consumption of resources, not even for the construction, because a solar
updraft tower predominantly consists of glass and cement which is sand plus self-made energy (Fig. 45)
- the (high) investment costs are almost exclusively due to labour costs. This creates jobs, and
- a high net product for the country with increased tax income and reduced socials costs, and in addition
- no costly imports of coal, oil, gas which is especially beneficial for the emerging nations such as China,
releasing means for their development.

October 2007 Page 25

ELECTRICITY FROM THE SUN SCHLAICH et all

Fig. 45: A solar updraft tower in the background feeding its energy into a cement plant (left) and a glass-
factory (right) which again supply the cement and glass for the construction of the next solar updraft
tower

8. PROSPECTS

Intensive design and testing of central and decentralized solar thermal power systems using different concepts is
currently underway. All technologies described before have already left the stage of prototype. They are poised
on the verge of commercialisation or have already reached that stage.
Still, for all these technologies it is currently true that the associated power generation costs – depending on labor
costs and interest rates - are higher than those for conventionally generated electricity. Thus, to further decrease
power generation costs, more development efforts are needed, but also the operation of power plants under real-
world conditions. The multitude of activities in the area of project development let expect that this goal will be
reached within the next few years. Taking the unavoidable rise of costs for the provision of fossil fuels in the mid
and long term into consideration, it becomes very probable that solar thermal power generation will gain ground
on the energy markets in the near future and that it will play an important role in the long run.
Beyond that we have no choice but to do something for the energy consent, the environment and above all for
the billions for underprivileged people in the Third World. But we should not offer them hand-outs, a multiple of
which we deceitfully regain by imposing a high interest rate on their debt. Instead we should opt for global job
sharing. If we buy solar energy from Third World countries, they can afford our products. A global energy
market with an essential solar contribution beyond hydropower is no utopian dream!

If we really want to we can do it!

October 2007 Page 26

 ELECTRICITY FROM THE SUN SCHLAICH et all

9. References (short selection)

[1] Schlaich, S. and J.: Erneuerbare Energien nutzen. Werner-Verlag, Düsseldorf, 1991
[2] Schlaich, J.: The Solar Chimney. Edition Axel Menges, Stuttgart, 1995
[3] Schlaich, J.: Renewable Energy Structures, Structural Engineering International, pp. 76 – 81, Vol. 4,
No. 2, 1994
[4] Stine, W.B.: A compendium of solar dish /Stirling technology, Sandia National Laboratories,
Albuquerque, NM and Livemeore, CA, SAND93-7026 UC-236, 1994
[5] Schiel, W. et al.: Long term testing of three 9 kW dish/Stirling systems. Proceedings of the ASME
International Solar energy Conference, 27 –30 March, San Francisco, CA, 1994
[6] Keck, T., Schiel, W.: Dish/Stirling-Anlagen zur dezentralen solaren Stromerzeugung. BWK: 53(12),
S. 60, 2001
[7] Keck u.a.: EnviroDish and EuroDish – system and status. Beitrag zum ISES (International Solar Energy
Society) Solar World Congress, Göteborg, Schweden, 2003
[8] Winter, C.J. et al. Hrsg. : Solar Power Plants, Springer, Berlin, Heidelberg, New York
[9] Weinrebe, G. : Technische, ökologische und ökonomische Analyse von solarthermischen Turmkraft-
werken, IER Forschungsbericht 68, Stuttgart, 2000
[10] Weinrebe, G. und Laing, D.: Solarthermische Stromerzeugung, In Kaltschmitt, M.; Wiese, A. (Edts.):
Erneuerbare Energien – Systemtechnik, Wirtschaftlichkeit, Umweltaspekte. 3. Auflage, Springer, Berlin,
Heidelberg, New York, 2003
[11] Keck et al.: Development and construction of a Metal-membrane Heliostat, Proceedings of the
6th International Symposium on Solar Thermal Concentrating Technologies, Mojacar, 1992.

October 2007 Page 27

View publication stats