GEOTHERMAL AND SOLAR THERMAL Professor Hal Gurgenci School of Mechanical and Mining Engineering The University of Queensland

CRICOS Provider No 00025B

How long to work for a chicken dinner

CRICOS Provider No 00025B

From Wikimedia Commons, the free media repository

Rice for less work but with more energy Rice Cultivation in Andhra Pradesh

CRICOS Provider No 00025B

Online at http://mpra.ub.uni-muenchen.de/49204/ MPRA Paper No. 49204, posted 21. August 2013 11:53 UTC

Electricity consumption per person

kWh/person

Australia

France

China World India Bangladesh

http://data.worldbank.org/indicator/EG.USE.ELEC.KH.PC/countries/1W-AU?display=graph

CRICOS Provider No 00025B

Lorentz and Gini Curves 100%

Gini coefficient is the ratio A/(A+Ao).

Fraction of the consumption

The area A is the yellow area. The area Ao is the area under the green (45o) line. A If the real world curve coincides with the perfect line, the Gini coefficient is zero. Ao 45o

For extreme inequality, the Gini coefficient is 1.

Extreme inequality 100% Fraction of the population

CRICOS Provider No 00025B

Energy Inequality

Question: World energy inequality was very high in 1980. It is decreasing since then. This means poorer countries are consuming more power. As this continues, we will need more and more energy sources. Where will the new energy come from? CRICOS Provider No 00025B

Max Roser -- http://ourworldindata.org/data/resources-energy/energy-production-and-changing-energy-sources/

Global Electricity Demand Projection

18000 TWh1 28000 TWh2

(1) (2)

CRICOS Provider No 00025B

http://data.worldbank.org/indicator/EG.USE.ELEC.KH.PC This is less than the BP projection. I am assuming the GFC slowdown will continue for a while taking the demand down.

Fossil Fuel Reserves CHINA Oil Coal Gas

Proven Annual Reserves Consumption Mtoe Mtoe Years 3238 167 19 58900 797 74 1641 28.9 57

250

How long will reserves last at the current consumption rates?

Years

200

150 World 100

China

50

H Gurgenci ([email protected])

WORLD Oil Coal Gas

Proven Annual Reserves Consumption Mtoe Mtoe Years 156700 3614 43 501172 2368 212 158198 2292 69

0 Oil

Coal

Gas

Source: Earth Trends Data Tables: Energy and Resources Original Source: Proved Fossil Fuel Reserves and Average Annual Fossil Fuel Production BP p.l.c., 2004. Statistical Review of World Energy. Available online at: http://www.bp.com/statisticalreview2004. BP compiles these statistics using a combination of primary official sources, third party data from the OPEC Secretariat, Cedigaz, World Energy Council, World Oil, Oil & Gas Journal, and an independent estimate of Russian reserves based on information in the 16 Oct 2008 public domain.

14

TWh

Fossil Fuel Reserves into the future

CRICOS Provider No 00025B

We will run out of fossil fuels in 120 years. The depletion date will be sooner if the energy-poor nations develop faster.

Energy Inequality is not sustainable • The energy poverty for the billions cannot be sustained. • Therefore, the future electricity generation growth may even be faster • I will address the first point, the unsustainability of energy poverty, on the next two slides.

CRICOS Provider No 00025B

Energy poor populations grow fast!

CRICOS Provider No 00025B

Future population growth

US Census estimates

World population estimates from 1800 to 2100, based on "high", "medium" and "low" United Nations projections in 2010 and US Census Bureau historical estimates. http://en.wikipedia.org/wiki/Projections_of_population_growth

CRICOS Provider No 00025B

Climate Change and other hazards • I hope I have been able to demonstrate that we have a serious problem before even starting to consider other issues such as – Air pollution due to fossil fuel – Global warming

CRICOS Provider No 00025B

Is cheap electricity worth the health risk?

AIR POLLUTION

CRICOS Provider No 00025B

What colour is the sky?

2015 LUOYANG

2012 BEIJING

CRICOS Provider No 00025B

The sky is blue away from the smokestacks

CRICOS Provider No 00025B

Yes, it will cost money. Is it worth it?

If the cost of this water were 1 cent, how much more would you be willing to pay for clean water? CRICOS Provider No 00025B

Most would be happy to pay five times as much or more to get this.

Fossil fuels and renewables

The electricity from these plants (you can hardly see them because of the pollution they make) costs 3 - 5 cents/kWh.

CRICOS Provider No 00025B

The electricity from the above plants costs 1015 cents/kWh now, and will get cheaper when we build more of them.

Electricity content of our income Country

kWh/person

GDP[$]/person

kWh/$ GDP

Australia

10000

43000

0.23

China

3500

12000

0.29

• Australians consume 0.23 kWh for every $ they make • This means Australians spend – 1.1% on electricity generation if using coal – 3% if they were using renewable energy

• Similar number for China – 1.4% if using coal – 4%if using renewable energy

• This means if we all switch to renewable energy now, we would be 2-3% poorer but we would have crystal clean air. CRICOS Provider No 00025B

The risk to the world as we know it

CLIMATE CHANGE

CRICOS Provider No 00025B

The world IS getting warmer -- slowly The term temperature anomaly means a departure from a reference value or long-term average.

https://www.ipcc.ch/pdf/assessment-report/ar5/syr/AR5_SYR_FINAL_SPM.pdf

CRICOS Provider No 00025B

some will change more than the others Figure 2 | Time series of the annual maximum wet bulb temperature(1) for each ensemble member and GHG scenario. Blue, green and red lines represent the historical (1976–2005), RCP4.5 (2071– 2100) and RCP8.5 (2071– 2100) scenarios, respectively. (1)

maximum daily value averaged over a 6-h window. Pal, J S, Elfatih, A B E (2015). Future temperature in southwest Asia projected to exceed a threshold for human adaptability. Nature Climate Change, DOI: 10.1038/NCLIMATE2833

WHY ? Carbon dioxide concentration in the air was reasonably stable before industrialisation. Since 1700, it has started increasing. The chart is based on measurements from Cape Grim and on air samples collected from Antarctic ice at Law Dome.

http://www.csiro.au/greenhouse-gases/

This is why we need new and clean energy sources. I will address only two: Geothermal and Solar Thermal.

CRICOS Provider No 00025B

The Earth is Hot Our drilling activities are limited to the top 5 km

Solid (5000-6000 oC)

The temperature increases as we go deeper into the crust. The average rate of increase is about 25 oC/km over the first 100 km. One would need a hot source of at least 150 oC for economic electricity generation using present technology. To get down to that depth, on average, one needs to drill 5 kilometers. 34

This is how hot the earth is

italy-sicily-stromboli-volcano-eruption.jpg http://www.zmescience.com/other/science-abc/types-of-volcano/ CRICOS Provider No 00025B

Krafla Geothermal Plant in Iceland

36

Nesjavellir Geothermal Power Plant in Iceland

37

Mutnovsk, Kamchatka, Russia

38

Old and New Plants, Tuscany

39

Wairakei, New Zealand

40

Mindanao, Philippines

41

North-East Africa

Ilegedi Bubbling Pool in Eritrea

Fumarole (steam vent) in Eritrea

42

Existing Geothermal Sites • • • • • • • •

California, USA Tuscany, Italy New Zealand Japan Iceland Kamchatcka, Russia Phillippines Eritrea

What is common about these locations?

43

World Volcanic Geothermal Resources

44

Enough geothermal energy to provide for the nation for the next 6000 years

Temperatures at 5 km 22,000 EJ

5000 MWe by 2030

Cost projections

$4-6m/MW

8-10 ¢/kWh

200 oC

45

A geothermal energy boom is expected once the energy economics change with pricing of carbon emissions This is a 2007 slide. This was what I predicted at that time. 46

47

Geodynamics, Innamincka, Central Australia

http://www.enviroinfo.com.au/wp-content/uploads/2013/05/Geothermal-SA.jpg

February 2014

H Gurgenci ([email protected])

All Power from the Sun ? This figure shows how large an area we need to meet the entire electricity demand for Europe and for the world.

CRICOS Provider No 00025B

Global Horizontal Irradiation Map http://solargis.info/doc/_pics/freemaps/1000px/ghi/SolarGIS-Solar-map-China-Mainlands-en.png

This is the total area we need to generate all of the present Australian electricity demand.

Total electricity generation = 250000 GWh/y DNI in Roma = 2500 kWh/m2/y Overall solar conversion efficiency = 25 % Solar electricity generation per m2= 625 kWh Mirrors required for Australia = 400 km2 CRICOS Provider No 00025B

The World Solar Resource ( GHI ) http://solargis.info/doc/_pics/freemaps/1000px/ghi/SolarGIS-Solar-map-World-map-en.png

GHI is relevant to PV

CRICOS Provider No 00025B

The World Solar Resource ( DNI ) http://solargis.info/doc/_pics/freemaps/1000px/dni/SolarGIS-Solar-map-DNI-World-map-en.png

DNI is relevant to Solar Thermal

CRICOS Provider No 00025B

Global Horizontal Irradiation

Diffuse Horizontal Irradiation – DHI

GHI is all solar incidence falling on the horizontal collector. Part of it direct from the sun, part of it diffuse.

Horizontal plate catches only PART of the DNI. CRICOS Provider No 00025B

Solar collecting surface (horizontal)

Direct Normal Incidence

Mirrors can only reflect the direct radiation

Solar concentrating mirror (tracking the sun) CRICOS Provider No 00025B

What Is Concentrating Solar Thermal?

Different uses for GHI and DNI • Photovoltaic systems are designed using GHI (usually adjusted for the panel tilt angle)

• Concentrating Solar Thermal systems are designed using DNI

CRICOS Provider No 00025B

DNI

Diffuse

Tilt Angle

PV ARRAYS at THE UNIVERSITY OF QUEENSLAND

March 2014 | Hal Gurgenci

CRICOS Provider No 00025B

HOLD ON A SECOND ! Hold on a second? What do we do when the sun goes out?

We use storage. Store during the day, live off the store at night.

Why does IEA predict CST to grow when PV is cheap and is getting cheaper ? 350

Generated using data in Figures 4.45 and 4.26 of Australian Energy Technology Assessment bree.gov.au (2012) 300

$/MWh

250

200

CST

150

100 PV 50

0 2010

2015

2020

2025

2030

2035

Years

2040

2045

2050

2055

2060

The question ! Why does IEA predict CST to grow when PV is cheap and is getting cheaper ?

The PV is intermittent. The CST is not.

UQ PV ARRAY ON A GOOD DAY

“intermittent” means no electricity when there is no sun !

the next day

HOLD ON A SECOND ! Hold on a second? If PV is intermittent, why is CST not? They both rely on the sun. When there is no sun, both will have to stop.

You beat intermittency if you can store energy. PV storage is expensive. CST is not.

Storing electricity is expensive Tesla’s PowerPACK for utilities unveiled in May 2015 has the price tag $250/kWh.

Cost of storing heat: $25/kWhth

$50/kWhe

Storing heat is cheap

PV or Solar Thermal ? We need both.

+ PV will produce electricity when there is sun.

CST will store heat when there is sun and produce electricity when there is NO sun.

PRESENT CST TECHNOLOGY IS NOT SUITABLE FOR THIS SCENARIO !

Concentrating Solar Thermal

CST PLANTS TODAY ARE NOT SMALL The PV + CST Hybrid Plants need small CST (1-30 MW). The present CST technology is based on steam power. Steam power cannot be made small. We need new technology.

The comparative advantages of supercritical CO2 over steam

Why Supercritical CO2? • Supercritical CO2 cycles are attractive for three fundamental reasons: • They are more efficient than steam cycle • They are more compact • They will be cheaper

More Efficient than steam Dostal, V., A Supercritical Carbon Dioxide Cycle for Next Generation Nuclear Reactors. 2004, MIT 60 He Brayton Supercritical steam Superheated steam sCO2

55

50

sCO2 is better than steam at T>560oC

Efficiency, %

45

40

35

This chart shows thermodynamic efficiencies of different cycle options. As you can see, supercritical CO2 cycle becomes better than supercritical steam at temperatures above 550 oC and it gets even better at higher temperatures.

30

25

20

400

500

600

700

Temperature, oC

800

900

More COMPACT than steam

Dostal, V., A Supercritical Carbon Dioxide Cycle for Next Generation Nuclear Reactors. 2004, MIT

STEAM POWER is COMPLEX This is the process sheet for a typical molten salt solar receiver with a steam power plant. Many turbine stages and many reheats are required to maintain the cycle efficiency.

the need for reheat Several reheats are needed to maximise the cycle efficiency in a steam power plant. All of this increases the complexity of the power plant. This is in addition to the steam turbines being bulkier than the supercritical CO2 turbines. RESULT: Steam power plants are not viable at sizes smaller than 50 megawatts. Supercritical CO2 powerplants can be built as small as 1-Mwe, possibly smaller.

Efficiency of small plants

Supercritical CO2 makes it possible to build small power plants with small fields. This may mean efficiency gains of around 5%.

M J Wagner, MSc Thesis, Simulation and Predictive Performance Modeling of Utility-Scale Central Receiver System Power Plants, 2008 – page 32, Figure 9

IVANPAH

~ 1000 m

The closest = 150 m The far end = 1000 m

2% loss on clear days; 5% loss on hazy days 10% loss on clear days; 25% loss on hazy days

CHEAPER ELECTRICITY

Solar Tower with a Supercritical CO2 Power Block

Figure 5.11, US Department of Energy, Sunshot Vision Study, 2012

SCO2 Advantage Summary • At 700oC, supercritical CO2 7% more efficient than supercritical steam. It will get much better at higher temperatures. •

A supercritical CO2 plant can be small (1 to 30 MWe), therefore it will enjoy higher solar collection efficiencies due to smaller solar fields (5%)

• A supercritical CO2 plant will be simpler and therefore will cost less • The end result will be cheaper electricity

market advantage for sCO2

• If available will achieve 100% market share very quickly. If this is so good, why are they still using steam?

Until recently, there was not much interest in CST. Large turbine manufacturers did not want to invest in sCO2 because the market was small.

What is stopping us ? Let us explore the problem and identify the technical challenges

CST PLANT using a sCO2 Cycle

REC

Simple supercritical CO2 cycle QH HTR

EXP

COM

Unfortunately, this would be a very inefficient plant.

CT

COM: Compressor HTR: Heater EXP: Expander CT: Cooling Tower QH: Heat input QC: Heat rejection

G

The simplest supercritical CO2 plant would have one heater to heat CO2, an expander to generate power, a cooling tower to reject heat, and a compressor to compress the cold CO2.

QC

We need a recuperator.

need a recuperator because … 600

C W

500

D

T, oC

400

Therefore, the simple configuration is not suitable for power generation. Therefore, the minimum configuration is where a recuperator transfers some of the heat from the hot CO2 to the cold CO2.

QH 300

QL

200

B

100

A 0 800

1000

1200

1400

1600

1800

2000

Entropy, J/kg-K

2200

2400

Only a small part of the heat input is converted into power during the expansion C-D. The rest of it is rejected.

2600

2800

What does a recuperator do? 600

500

T, oC

400

300

C

With a recuperator, the bulk of the heat input before the turbine inlet comes from the CO2 stream still hot after the turbine exit. This is shown in the chart. The shaded area is the recuperator heat transfer. Both the heat input, QH, and the heat dump, QL, are reduced significantly. The cycle efficiency is doubled when a recuperator is used.

W

QH

D

QL

200

B 100

0 800

QR

A 1000

1200

1400

1600

1800

2000

Entropy, J/kg-K

2200

2400

2600

2800

RECUPERATED CYCLE The recuperator design is of critical importance. It should have low pressure drop and high heat transfer coefficient. Heatric Printed Circuit Heat Exchangers offer the only proven option at the moment.

https://mercureaace2013.wordpress.com/2013/03/28/w1 0_aib_replacing-printed-circuit-heat-exchanger/

Increased efficiency with increasing complexity 3

4 6

R

5

R

0

9 8 Using two compressors and two recuperators increase the efficiency a bit more and this might justify the extra cost.

G

E

C

C

7 1

2

what is available already? • Solar Mirrors • Solar Tower and the Receiver • Thermal Storage • Compressor • Recuperator • Cooling Tower • Turbine • Generator

Component options Turbine

Missing technology. No commercial product yet.

Type How hard can it be to • Axial for utility-scale build a supercritical • Radial for small systems CO2 turbine? Compact size • High RPM higher power density than steam • Low pressure ratio fewer stages than steam

Compressor

Commercial products exist

Available for pipeline CO2 transport and enhanced oil recovery (EOR) applications.

Recuperator

Commercial products exist

Printed Circuit Heat Exchangers used in gas-to-gas heat transfer in process industries. sCO2 is denser than most gases smaller exchanger for the same heat transfer duty compared to process gases

Heater/Thermal Storage

Several options exist.

High pressure CO2 in standard tubes. The type of thermal energy storage dictates what is outside.

Cooling Tower

Commercial products exist

Hybrid cooling is preferred (commercial air cooling enhanced using occasional evaporative cooling)

Near the Critical point SC Steam

SC CO2

560

560

Pressure, kPa

28000

20000

Density, kg/m3

88

123

Turbine Inlet Temperature, oC

Cooling Tower Temperature, oC

30

Pressure, kPa

4

Density, kg/m3

0.03

98

30

9000 995

166

746

High risk high reward

THOSE WHO DARE WIN !

2013 Total Electricity Production in China (ex. Hong Kong) = 5400 TWh

FUture SOLAR THERMAL Growth

Source: Figure 10, International Energy Agency, Technology Map – Solar Thermal, 2014 Edition

MARKET Growth 1200

The plot generated using the data on the previous slide and a unit cost of $7m/MWe.

CST Unit Cost = $7m/MWe Breakdown from Figure 4.2, IRENA, CSP Cost Analysis, 2012

Total GWe and $b/year

1000

800

Global installed CST capacity, GWe

600

400

200

Annual investment, $b 0 2010

2015

2020

2025

2030

Years

2035

2040

2045

2050

sCO2 turbine/compressor market (assuming 100% CST market penetration) sCO2 turbine/compressor spending, $ billions

250

200

150

100

50

0 2015

2020

2025

2030

Years

2035

2040

2045

2050

SUPERCRITICAL CO2 SYSTEM AND TURBINE DEVELOPMENT AT THE UNIVERSITY OF QUEENSLAND

Turbine design challenges • The challenges are • Aerodynamic design for the rotor and the nozzles • Mechanical design of the turbine • Bearings • Seals • Generator

AERODYNAMIC DESIGN • Turbine Design Process 1. Meanline Design – –

2D geometry e.g.Inlet/Outlet radius Performance estimation e.g. power and efficiency

2. CFD Simulation – –

3D geometry e.g. thickness and angle distribution of the curved rotor blade Performance estimation e.g. power, efficiency and chocking

PhD Student : Mostafa Odabaee

Mechanical Analysis The key characteristics of sCO2 turbine can include; – – –

30mm rotor for a 500-kWe turbine

High density gradient, High pressure working fluid and, Small size of turbine.

Impeller blades in sCO2 radial inflow turbines are exposed to; – –

High thermal loads and centrifugal forces and, Additional dynamic stresses occur by the aerodynamic excitation.

This is one of the early designs being considered.

FE analysis; – Deformation and stress can be calculated using pressure and temperature loads from CFD Vibration analysis; November 2014 December 2014 One-way coupling FSI is chosen based on following procedure; 1. The unsteady pressure field on the entire rotor surface are extracted using transient CFD. 2. A Fourier decomposition is conducted for the unsteady pressure field at any location of the CFD grid. 3. The harmonic forced response then is computed, applying the pressure forces from CFD.

PhD Student : Mohsen Modirshaneschi

SEALS i.

no current commercial dry gas seal is available for operating under S-CO2 conditions therefore a functional new design operating under S-CO2 condition is needed detail studies is needed in order to properly design a S-CO2 . ii. old empirical methods based on incompressible fluid or ideal gas are not working for real gas S-CO2 properties abrupt changes in a non-linear fluid properties could introduce some interesting flow behaviour such as sonic effect etc. if not properly overcome these issue might not work. GROOVE

DIRECTION OF ROTATION

PhD Student : Mohd Fairuz Zakariya

Grooves

Bearings Background • • • •

Oil-free system[1] Increase tolerance with respect to deformations of rotor/stator Better tolerances to manufacturing variation Better rotor dynamics

Research Methodology Code development • Fluid domain simulations: Eilmer • Conjugate heat transfer (CHT) analysis: Eilmer & OpenFOAM • Structural deformation: written in Python • Moving grid development in CFD code: Eilmer • A full 3D code: Eilmer & OpenFOAM & Python Validation with experiments PhD Student : Qin Kan

UQ sCO2 Turbine test facility

TEST LOOP diagram

uq turbine prototype This is a refrigerant turbine. It was easier to design and build but we re hoping to validate our design software by testing this prototype. When validated, the same software can be used to design a supercritical CO2 turbine. We are hoping to design and test our first supercritical CO2 turbine in 2016.

The rotor

The test set-up

TEST SET-up Cut-away drawing

UQ PROJECTS Exists

Exists A

Proposed

C

B