Mounted Small-Scale Solar Thermal Brayton Cycle

THERMODYNAMIC OPTIMISATION AND EXPERIMENTAL COLLECTOR OF A DISH- MOUNTED SMALL-SCALE SOLAR THERMAL BRAYTON CYCLE

WG LE ROUX

Study-leaders: Prof. T. Bello-Ochende Prof. J.P. Meyer Department of Mechanical and Aeronautical Engineering, University of Pretoria, South Africa

February, 2015

Submitted in partial fulfilment of the requirements for the degree PhD (Mechanical Engineering)

1 Presentation Outline

1. Introduction 2. Background 3. Literature Study 4. Modelling and Optimisation 5. Analytical Results 6. Experimental Study 7. Conclusion 8. Recommendations

2 1. Introduction

Parabolic dish concentrator for a Stirling engine (Image extracted from Pitz-Paal, 2007)

Long-term average of direct normal solar irradiance on a world map showing the potential of solar power generation in southern Africa (GeoModel Solar, 2014)

A typical micro-turbine (the GT1241) as available from Honeywell, Garrett proposed for the small-scale solar thermal Brayton cycle (Image extracted from Garrett, 2014) 3 1. Introduction

Problem • Solar-to-electricity technologies are required which are • more competitive • more efficient • cost-effective

Purpose of the study Small-scale dish-mounted open solar thermal Brayton cycle • optimise solar receiver and recuperator - method of total entropy generation minimisation • test optimised receiver

Objectives • Second law of thermodynamics • Entropy generation minimisation • Ray-tracing software • Geometry optimisation • Experimental receiver test

4 2. Background

Scope of Research – Thermodynamic Optimisation

• Open and direct solar thermal Brayton cycle • Second Law of Thermodynamics • Entropy Generation Minimisation • Maximise net power output • Optimise geometry of recuperator and receiver • Heat Transfer & Fluid Flow Irreversibilities • Experimental setup Solar resource – South Africa Why Solar?

Solar resource - World

• According to DLR Solar resource – South Africa Why Solar?

The Department of Minerals and Energy places South Africa’s annual direct normal irradiation (DNI) between 2 500kWh/m2 and 2 900 kWh/m2 with an average of almost 300 days of sunshine per year.

Solar resource – South Africa, Pretoria Meteonorm

1400

1200

1000

800

600 Irradiance of beam

Mean irradiance of global

Irradiance (W/m^2) Irradiance 400 radiation, tracked Mean irradiance of global radiation horizontal 200

0 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Time (h) CSP - Concentrating methods

Dish Trough Tower Background

Small-scale solar power Brayton cycle • Photovoltaic cells • mobility, cost benefits • Solar water heaters • micro-turbines • CSP (Concentrated solar power) • hybrid system – Trough • storage systems • Rankine Cycle • water heating • efficient and highly competitive. – Dish-mounted • Stirling Engine Maximum net power output • Brayton cycle • combined effort of • heat transfer, Air out • fluid mechanics and 11

• thermodynamics 10 9

Recuperator 3 4    8 Wnet Wt Wc 2

Compressor Turbine Load

Receiver 5 7 6  1 Air in Q* 12

Solar tracking - Elevation

• SunEarthtools Solar tracking - Azimuth

Measured angle of tracking system versus real azimuth angle of the sun 150 Morning measurements 100 Noon measurements Afternoon measurements 50 SunEarthTools

Azimuth Azimuth angle -50

-100

-150 6 7 8 9 10 11 12 13 14 15 16 17 18 Time (h) Two-axis solar tracking required for dish Solar Active tracking

Micro- Date and time Passive processor based or a and combination Bi- Fluid electro- of sensor and Auxiliary optical date/time metallic bifacial strips sensor based solar cell based based

Mousazadeh et al. (2004), Poulek and Libra (2000) 3. Literature Study

Air out 11

10 9

Recuperator 3 4    8 Wnet Wt Wc 2

Compressor Turbine Load

Receiver 5 7 6 1 Air in Q *

The open and direct solar thermal Brayton cycle

16 3. Literature Study

Test set-up of a solar thermal Brayton cycle (Image extracted from Heller et al., 2006)

17 Small-scale open and direct solar thermal Brayton cycle with recuperator • Advantages – High recommendation – Air as working fluid – Hot air exhaust • Water heating • Space heating • Absorpsion refrigeration – Recuperator • high efficiency and • low compressor pressure ratios

• Disadvantages – recuperator and receiver pressure losses – turbo-machine efficiencies – recuperator effectiveness – Heat losses irreversibilities Solar thermal Brayton - Recuperator

Air out 11

10 9

Recuperator 3 4    8 Wnet Wt Wc 2

Compressor Turbine Load

Receiver 5 7 6 1 Air in Q *

Solar thermal Brayton - Recuperator

Image extracted from: Stine, B.S., Harrigan, R.W., 1985, Solar energy fundamentals and design. New York: John Wiley & Sons, Inc. 3. Literature Study

Efficiencies of different solar receivers – Pressurised volumetric

Receiver Reference T (K) T (K) P (kPa) (kg/s) Working fluid ΔP (Pa)  out in type number or rec m model Pressurised PLVCR-5 71% 1 323 - 420 - Air - volumetric (Ávila-Marín, 2011) PLVCR-500 57% 1 233 300 415 - Air - (Ávila-Marín, 2011) DIAPR 79% 1 477 308 1 800 0.0222 Air 25 000 (Karni et al., 1997), (Ávila-Marín, 2011) REFOS 67% 1 073 - 1 500 - Air 1 800 (Buck et al. 2002), (Ávila-Marín, 2011) Dickey, 2011 88% 871 542 273 0.409 Air 2 900 21 3. Literature Study

Efficiencies of different solar receivers - Tubular

Receiver Reference Tout (K) Tin (K) P (kPa) (kg/s) Working fluid ΔP (Pa) type number or  rec m model Tubular Cameron et 51%* 1 089 865 370 0.73 He-Xe 7 000 al., 1972 Kribus et al., - 1 023 300 1 600 - 0.01 Air 40 000 1999 1 900 Heller et al., - 823 573 650 - Air 10 000 2006 Neber and 82% 1 500** - 760 0.0093 Air 40 Lee, 2012 Amsbeck et 43% 1 076 876 384 0.526 Air 7 330 al., 2010 Amsbeck et 39.7% 1 055 871 375 0.516 Air 7 400 al., 2010 Solugas - 873 598 850 5.6 Air (Quero et al., 2013) *calculated by author **proposed

22 3. Literature Study

Particle receiver Open volumetric receiver – HiTRec (Image extracted from Miller and Koenigsdorff, (Image extracted from Ávila-Marín, 2011) 1991)

23 3. Literature Study

Closed volumetric receiver, Longitudinal tubular receiver REFOS (Image extracted from Amsbeck et al., 2008) (Image extracted from Buck et al., 2002)

24 3. Literature Study

Ceramic counterflow plate-type recuperator Coiled tubular receiver (Image extracted from Pietsch and Brandes, 1989) (Image extracted from Kribus et al., 1999)

25 3. Literature Study

Q1 = 6.8 kW, T1 = 1 308 K, Q2 = 8.3 kW, T2 = 1 179 K, Q3 = 9.7 kW, T3 = 1 054 K, Q4 = 11.2 kW, T4 = 904 K Q5 = 12.7 kW, Q6 = 14.1 kW, Q7 = 15.9 kW

Performance map 12

(in different weather conditions) Q6 10 Q • small-scale open solar thermal 7 8 Brayton cycle Q5 T1

• fixed optimised geometries (kW) 6 T2

T3 4

T4 2 Q4 Q Q2 1 Q3 0 1.4 1.6 1.8 2 2.2 2.4

26 4. Modelling and Optimisation

Control volume for the open solar thermal Brayton cycle

 Qloss, j j

m    m Wnet Wt Wc

Q *

27 4. Modelling and Optimisation

Solar receiver - SolTrace

Example of an analysis done for the solar dish and receiver

28 4. Modelling and Optimisation

Solar receiver

Rectangular open-cavity Heat loss from the solar receiver open-cavity receiver

29 4. Modelling Solar receiver air heating • Rectangular open cavity tubular receiver • Stainless steel • Pressure drop (Colebrook equation)

Variables • Tube diameter, • Inlet temperature, • Mass flow rate

30 4. Modelling Solar receiver – conduction heat loss [1]

Assumptions: • Wind speed: 2.5 m/s

• T0 = 300 K • P0 = 86.6 kPa • 100 mm insulation thickness • Conductivity of 0.061 W/mK at 550 °C average temperature [2] • Elevation angle of 45 °

(1/ hout tins / kins) 1.77 A T  T T  T  n  s,n    s,n   Qloss,cond,n   Rcond 1/ hout An  tins / kins An  [1] Le Roux, W.G., Bello-Ochende, T. and Meyer, J.P., 2014, The efficiency of an open cavity solar receiver for a small-scale solar thermal Brayton cycle, Energy Conversion and Management 2014;84:457–70. 31 [2] Harris, J.A., Lenz, T.G., 1983, Thermal performance of solar concentrator/cavity receiver systems, Solar Energy 34 (2), pp. 135-142.

4. Modelling Solar receiver – radiation heat loss

 4 4 Qloss,n,rad   AapTs,n T  N  4 4 Qn  An  Fn j  nTs,n   jTs, j  j1

32 4. Modelling Solar receiver – convection heat loss [2]

Koenig and Marvin heat loss model [2]  Qloss,conv,n  whinner An Ts,n T 

2a2 Nu  h cav inner  3.2 1/ 4  0.52(cos )GrL Pr

For aopt = 0.25 m [1]: h = 2.75 W/m2K [1] Le Roux, W.G., Bello-Ochende, T. and Meyer, J.P., 2014, The efficiency of an open cavity inner solar receiver for a small-scale solar thermal Brayton cycle, Energy Conversion and w = 2 Management 2014;84:457–70.

[2] Harris, J.A., Lenz, T.G., 1983, Thermal performance of solar concentrator/cavity receiver systems, Solar Energy 34 (2), pp. 135-142. 33 4. Modelling and Optimisation

Recuperator design in SolidWorks Recuperator geometry

34 4. Modelling Recuperator

t H

Lreg

• Counterflow plate-type recuperator • Pressure drop : Colebrook equation • Fully developed laminar flow • t = 1 mm • Geometry variables: a, b, L, n 35 4. Modelling Recuperator UA Efficiency modelling: NTU h  Updated version of the ε-NTU – method [3] m h c p0,h • Includes heat loss to the environment m h c p0,h • Since recuperator operates at high temperature Crh  m c c p0,c 1  1  ,Cr 1  X 0   X 1 h   ,Crh 1      Cr Qloss,c h   c  h    Crh 1  X 1 ,Crh 1   c 1  X 0 ,Crh 1 UATh,in Tc,in   Qloss,h h  E UATh,in  Tc,in  B   h  Crh  c 1 e   X 0   1  Cr 1 e E    h    Cr 1  Crh   h  B   NTUh c  h  Crh 1  Crh   X 0 1  X 1  NTUh c  h  1 Crh E  NTUh Crh 1 [3] Nellis, G.F. and Pfotenhauer, J.M., 2005, Effectiveness-NTU relationship for a counterflow heat exchanger subjected to an external heat transfer, Journal of Heat Transfer 127, pp. 1071 – 1073. 36 4. Modelling Micro-turbine

Standard off-the-shelf micro-turbines from Honeywell • Geometry not optimised • Compressor map • Isentropic efficiency • Corrected mass flow rate • Pressure ratio • Rotational speed

37 4. Modelling Micro-turbine

Standard off-the-shelf micro- turbine from Honeywell • Parameter: turbine operating point • Turbine map • Corrected mass flow rate • Pressure ratio • Maximum efficiency • Efficiency as function of 2N  D   t  pressure ratio found using 60  2  BSR  2 1/ 2 blade speed ratio (BSR)   BSR  0.6     1k    1    2h 1 r k t t,max    in  t    0.6      38 4. Modelling Receiver heat flux • Receiver heat flux determined with SolTrace • Solar tracking error of 1° • Optical error of 10 mrad • Dish reflectivity of 85% • Direct normal irradiance of 1 000 W/m2

39 4. Modelling Net absorbed heat rate

Determined for each tube section

 n1         Qnet,i  Qnet,n  Qsolar,n  Qloss,rad,n  Qloss,conv,n  Qloss,cond,n T    T   s,n    in,0  N i1  mc p0    4 4    Qnet,n  Qsolar,n  An Fn j  nTs,n  jTs, j  Qnet,n    1 1  j1    4 4     An Fn  nTs,n  jT   hAn 2mc p0   hn An Ts,n T  An Ts,n T / Rcond • Equations are solved Q  Q  A   m T  c  simultaneously with net,n solar,n n n 1 s,n 1 N Gaussian elimination 4  An  Fn j  j m1Ts, j  c1  An nFnT • Radiation heat loss term j1 is linearised An  An m2Ts,n  c2  Ts,n T  Rcond 40

4. Modelling Net absorbed heat rate

Determined for each tube section

1400

1200

1000

800

600 • Equations are solved simultaneously with 400 Gaussian elimination • Radiation heat loss term 200 is linearised 0 0 5 10 15 20 25 Tube position - bottom to top 41

4. Modelling Net power output Objective function:  T   T         1  Wnet  TSgen,int  1 Q*mcp0 T1 T11 mTcp0 ln   T * T11 

S   m c ln T / T  m R ln P / P gen,int  p0  1 2   1 2 compressor m  P /14.7 tCF 7  Q / T  m c lnT / T  m R lnP / P  m t   l  p0 3 2 3 2 Duct23 T7  460/ 519 R / c  T T  P P  p 0    m c ln 10 4  10 4    Q / T  • Steady-state  p0 T T  P P  l     9 3  9 3    temperatures and recuperator  Q / T  m c ln T / T  m R ln P / P pressures found with  l  p0  5 4   5 4 Duct45 iteration, written in    Q* Qloss  terms of isentropic     m c p0 lnT6 / T5  m R lnP6 / P5  T * T efficiencies,    receiver recuperator efficiency,   Ql / T  m c p0 lnT7 / T6  m R lnP7 / P6  geometry variables Duct67   m c ln T / T  m R ln P / P  p0  7 8   7 8 turbine  Q / T  m c ln T / T  m R ln P / P 42  l  p0  9 8   9 8 Duct89 4. Modelling Net power output

Air out 11

10 9 Assumptions: Recuperator 3 4    • Connecting tubes 8 Wnet Wt Wc 2 • Insulation Compressor Turbine Load • 0.18 W/mK conductivity Receiver 5 7 • 10 mm thick 6 • T8 = T9, T2 = T3 1 Air in Q * • P8 = P9, P2 = P3 • V1 = V11 • Z1 = Z11 • Pressure drop – Colebrook equation (rough stainless steel friction factor)

• T1 = 300 K • P1 = P10 = P11 = 86 kPa • Steady-state temperatures and pressures found with iteration, using isentropic efficiencies, recuperator efficiency

4. Modelling Net power output

Air out 11

Optimisation: 10 9

Run through all different Recuperator 3 4 combinations of receiver    8 Wnet Wt Wc diameters, recuperator 2 geometries, micro- Compressor Turbine Load Receiver turbines and micro-turbine 5 7 operating points 6 1 Air in Q * MATLAB: For 3 different receiver tube diameters For 5 different micro-turbines For the different operating points of the turbine For 625 different recuperator geometries Find temperatures and pressures in the cycle with iteration Determine net power output

4. Modelling Constraints

• Maximum receiver surface temperature • 1200 K

• Recuperator total plate mass • 300 kg • 400 kg • 500 kg

45 4. Modelling - Flownex

Flownex modelling of the small-scale solar thermal Brayton cycle.

46 5. Analytical Results

Optical efficiency of a solar dish and receiver with a tracking error of 1° • SolTrace

0.9

0.8

0.7

0.6

0.5

0.4

0.3 Optical error = 5 mrad Optical error = 20 mrad 0.2 Optical error = 35 mrad 0.1 Optical error = 50 mrad 0 0 0.01 0.02 0.03 0.04 0.05 0.06 A’ =Area ratio (Aaperture/Aconcentrator)

47 5. Analytical Results

Overall receiver efficiency for a solar tracking error of 1° with receiver surface emissivity of 0.7 • Heat loss

0.9 Optical error = 5 mrad 0.8 Optical error = 10 mrad Optical error = 15 mrad 0.7 Optical error = 20 mrad Optical error = 35 mrad 0.6

0.5

0.4

0.3

0.2

0.1

0 0 0.002 0.004 0.006 0.008 0.01 A’ =Area ratio (Aaperture/Aconcentrator)

48 5. Analytical Results

Heat flux rate at different positions on the different receiver inner walls for a tracking error of 1° • SolTrace 100 Top 90 Side1 80 Opposite Side1 Side2 70 Opposite Side 2 60

Heat flux flux Heat(kW/m^2) 30

0 1 2 3 4 5 6 7 8 9 10 Position - bottom to top

49 5. Analytical Results

Temperatures and net heat transfer rates for a 0.0833 m receiver tube diameter with a tracking error of 1° and optical error of 10 mrad.

1400

1200

1000

800

600

400

200

0 0 5 10 15 20 25 Tube position - bottom to top

50 5. Analytical Results

Maximum net power output of the solar thermal Brayton cycle with a micro-turbine selected from Garrett 2.5 D = 0.0833 D = 0.0625 2 D = 0.05

rt = Turbine pressure ratio 1.5 D = Receiver tube diameter (m)

Wnet (kW) 1

0.5

0 1.2 1.4 1.6 1.8 2 2.2 rt

51 5. Analytical Results

Predicted temperatures at different positions in the solar thermal Brayton cycle • Matlab model • Flownex model

1200

Flownex Air out 1000 Matlab 11

10 9 800 Recuperator 3 4    8 Wnet Wt Wc

600 2 T T (K)

Compressor Turbine Load 400 Receiver 5 7 6 200 1 Air in Q *

0 Micro-turbine GT2560R at 87 000 rpm 1 2 3 4 5 6 7 8 9 10 Position in the cycle

52 6. Experimental Study

53 6. Experimental Study Solar dish and tracking system

Assembly of 4.8 m diameter parabolic solar dish in the laboratory (upside down):

Test set-up showing solar dish on two-axis solar tracking system: • SolidWorks • As constructed for experiment

54 6. Experimental Study Solar dish and tracking system

55 6. Experimental Study

Measured error of the end-height of the Absolute slope error per dish arm as 12 dish arms during pre-assembly and installed on the solar tracking on the tracker: system:

10 12

5 10

0 8 1 2 3 4 5 6 7 8 9 10 11 12

-5 6 Error Error (mm)

-10 4 Slope error error (mrad) Slope

-15 2 Pre-assembly On tracker -20 0 Segment number 1 2 3 4 5 6 7 8 9 10 11 12 Segment number

56 6. Experimental Study

Measurement of the solar resource Solar measuring station to measure the DNI of the sun (SOLYS 2): • Roof of Engineering Building 1

57 6. Experimental Study

Solar receiver

Manufacturing of solar receiver

58 6. Experimental Study

Solar receiver

Manufacturing of solar receiver

Inlet

Outlet

Side view of solar receiver

Position of three weldpad themocouples

59 6. Experimental Study

Solar receiver

Top view of the solar receiver with aperture shown at the bottom.

On the insulation before installation

60 6. Experimental Study

3 Test A – With blower Test B – Without blower 4

Layout of the experimental set-up. 1– Solar receiver with insulation; 2 – Leaf blower at receiver inlet; 3 – Receiver support structure; 4 – Parabolic dish; 5 – Thermocouple wires to data logger.

61 6. Experimental Study

A bottom view of the solar receiver and its support structure

62 6. Experimental study - results

Test A - Receiver surface temperature and air temperature measurements at the inlet (bottom) and outlet (top)

C) °

Temperature ( 20 Top Bottom 10 Air out Air in 0 0 200 400 600 800 1000 1200 1400 Time (s)

63 6. Experimental study

Test A - Steady-state receiver surface temperature and air temperature measurements at the inlet, outlet and in the middle of the receiver

Day 1 1 2 2 2 3 3 3 3 Blower setting 6 1 5 4 3 2 4 4 3 Start time 12:37 14:36 11:18 12:27 14:26 10:13 11:24 12:25 14:28 Steady-state time 13:00 14:56 11:34 12:52 14:40 10:41 12:01 12:45 14:41 Receiver inlet (°C) 39.2 38.8 35.5 38.4 35.9 35.5 - 38.0 36.2 Receiver middle (°C) 45.5 44.9 41.7 45.7 44.4 44.6 - 46.6 45.0 Receiver outlet (°C) 50.4 50.6 46 54.1 50.0 50.1 - 52.2 48.0 Air ambient (°C) 19.8 20.4 17 16.4 18.6 15.9 18.4 19.1 19.9 Air outlet (°C) 52 51 42 49 49 46 50 52 45.0 Collector efficiency 29.5 23.2 19.9 24.9 22.4 21.0 25.3 26.3 21.2 (%) Optical efficiency (%) 53.6 42.2 36.2 45.3 40.7 38.2 46.0 47.8 38.5

64 6. Experimental study

Test A - Expected ray performance of the experimental collector during the second test of Day 2, according to SolTrace.

For a dish with • 5 mrad slope error, • 25 mrad specularity error, • 1° tracking error, • 55% dish reflectivity, • DNI of 700 W/m2 and • 85% receiver tube absorptivity.

According to SolTrace, such a collector would have an efficiency of 21%.

This efficiency compares well with the efficiency of 23.2% obtained experimentally during the second test on Day 1 when the DNI was 700 W/m2.

65 6. Experimental study

Test B - Receiver surface temperature increase as a function of time • No blower 673 Top 623 Middle Bottom Insulation 573 Air

523

473

423 Temperature (K)

373

323

273 0 5000 10000 15000 20000 Time (s)

66 6. Experimental study

Test B - Receiver average surface temperature as a function of time • as measured experimentally • as calculated with • h = 6.5 W/ m2K before steady state • and h = 1 W/m2K after steady state

623 Measured 573 Calculated

523

473

423 Temperature (K)

373

323

273 0 5000 10000 15000 20000 Time (s) 67 6. Experimental study

Test B - Conduction heat loss from the receiver • as measured experimentally • as calculated with • h = 6.5 W/ m2K before steady state • and h = 1 W/m2K after steady state

300

Measured 250 Calculated

200

150 Heat loss Heat loss (W) 100

0 0 5000 10000 15000 20000 Time (s) 68 6. Experimental study

Test B - Receiver insulation change Heat loss from the receiver at an average temperature of 590 K with different insulation arrangements

100 90 Conduction 80 Convection 1 70 Radiation 60 50 40 30 2 20

10 Percentage of total heat loss (%) 0 1 2 Test Number

69 6. Experimental study

Receiver insulation change

Receiver surface temperature rise after insulation change

700

600

500

400

300

Temperature (K) Top 200 Middle

100 Bottom

0 0 2000 4000 6000 8000 10000 12000 Time (s)

70 7. Conclusion

• The method of total entropy generation minimisation was found to be a holistic optimisation approach whereby the components of the small-scale solar thermal Brayton cycle could be optimised.

• A method to determine the surface temperatures and net heat transfer rates along the length of the open-cavity receiver tube was presented.

• The factors contributing to the temperature and net heat transfer rate profiles on the receiver tube were divided into two components: • geometry-dependent and • temperature-dependent.

• It was found that many errors existed due to the solar collector – modelled with SolTrace

• An optimum receiver-to-concentrator-area ratio of A’ ≈ 0.0035 • for 1° solar tracking error, • 10 mrad optical error and • 45° rim angle was found for the open-cavity tubular solar receiver.

71 7. Conclusion

• The open-cavity tubular solar receiver surface temperature and net heat transfer rate for heating air depended on • the receiver size, • mass flow rate through the receiver, • receiver tube diameter, • receiver inlet temperature and • dish errors.

• Receiver efficiencies of between 43% and 70% were found for the open- cavity tubular receiver • with a = 0.25 m, • 0.06 kg/s ≤ mass flow rate ≤ 0.08 kg/s, • 0.05 m ≤ d ≤ 0.0833 m and

• 900 K ≤ Tin,0 ≤ 1 070 K, • operating on a 4.8 m diameter dish with 10 mrad optical error and maximum solar tracking error of 1°.

72 7. Conclusion

• The higher the mass flow rate through the receiver, the lower the surface temperatures and the more efficient the receiver.

• A high receiver efficiency was not necessarily beneficial for the small- scale solar thermal Brayton cycle as a whole but the second law efficiency was more important.

• The small-scale open solar thermal Brayton cycle could generate a positive net power output with solar-to-mechanical efficiencies in the range of 10-20% with much room for improvement.

• Optimum receiver and recuperator geometries were found.

• Good comparison between the Matlab results and Flownex results were found (within 8%), except for the recuperator outlet temperature, which differed because of the use of different ε-NTU methods to calculate the recuperator efficiency.

73 7. Conclusion

• A 4.8 m parabolic aluminium dish with rim angle of 45° and two-axis tracking system was designed and built.

• A tubular stainless steel solar cavity receiver was built and tested. • The efficiency of the collector was determined with a flow test. • A high-temperature test was performed to validate heat loss models. • The higher the inlet temperature, the less efficient the receiver became and the higher the maximum receiver surface temperature. • The convection heat transfer coefficient was determined • The heat loss rate due to convection and conduction was significantly reduced with the proper insulation arrangement.

• The use of SolTrace was validated to a certain extent.

• It is concluded that the small-scale dish-mounted open solar thermal Brayton cycle with tubular receiver and recuperator does have merit and it is recommended that it be investigated further experimentally.

74 8. Recommendations

• To make the small-scale open solar thermal Brayton cycle a success: • large receiver tube diameter, • very precise solar tracking system, • high-specularity, high-reflectivity dish, • 1° tracking error and 10 mrad optical error with reflectivity above 90% should be sufficient

• Future work • A smaller, more accurate and efficient dish and tracking system • Testing of the optimised open-cavity tubular receiver at a temperature of 1 150 K for fatigue loadings and thermal expansion • The optimised receiver should be coupled to an optimised recuperator and micro-turbine to determine the net power output of the system experimentally • A cost-effective high-temperature and low-emissivity stainless steel receiver coating should be developed. • Optimisation of the cycle at receiver surface temperatures below 700 °C so that black chromium can be used as low-emissivity coating.

• A moulded receiver cover to insulate the receiver • so that air cannot flow around the receiver tubes but only on the inner side of the receiver cavity • good thermal contact between the insulation and the receiver should be achieved regardless of thermal expansion • thermal expansion of the receiver should be considered

75 Acknowledgements

Assistance while building the fairly large experimental set-up: • Chris Govinder, • Prof Bello-Ochende • Donald Keetse, • Prof Meyer • Evan Huisamen, • Rupert Stander, • I thank my wife and my family for their • Koos Mthombeni, support. • Clyde Engineering, • Marcelino Benjamin, • Matsemela Zacharia (Zakes) This work is based on the research supported • Mogashoa, Milton Griffiths, • Otto Scheffler, by the National Research Foundation (NRF), • Ruan Fondse, University of Pretoria, CRSES, the Solar • Wian van den Bergh, • Johannes Joubert, Hub between the University of Pretoria and • Andries Tiggelman, Stellenbosch University, TESP, NAC, • Bera Chirwa, • Ryan Capitani, EEDSH Hub, Energy-IRT and the CSIR. The • Suzanne Roberts, financial assistance of the National Research • Jacob Masingi, • Milga Manufacturing, Foundation (NRF) towards this research is • Werner Scholtz, hereby acknowledged. Opinions expressed • Phenyo Zobane, • Erick Putter, and conclusions arrived at are those of the • Edwyn Mothabine, author and are not necessarily to be • Alan Naidoo, • Tebogo Mashego, attributed to the NRF. • Johan Clarke, • Modupe Matolo, • Israel Mabuda, • Thato Mahlatji, • James Gerber 76 • Zimase Dlamini.

I thank God for good health and an injury-free research period.

77 Journal Publications

1. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2011. Operating conditions of an open and direct solar thermal Brayton cycle with optimised cavity receiver and recuperator. Energy, Vol. 36, pp. 6027-6036.

2. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2012. Thermodynamic optimisation of the integrated design of a small-scale solar thermal Brayton cycle. International Journal of Energy Research, Vol. 36, pp. 1088-1104.

3. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2012. Optimum performance of the small- scale open and direct solar thermal Brayton cycle at various environmental conditions and constraints. Energy, Vol. 46, pp. 42-50.

4. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2013. A review on the thermodynamic optimisation and modelling of the solar thermal Brayton cycle. Renewable and Sustainable Energy Reviews, Vol. 28, pp. 677-690.

5. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2014. The efficiency of an open-cavity solar receiver for a small-scale solar thermal Brayton cycle. Energy Conversion and Management, Vol. 84, pp. 457-470.

78 Conference papers

1. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2011. Optimum performance of the small-scale open and direct solar thermal Brayton cycle at various environmental conditions and constraints. In: Proceedings of the International Green Energy Conference (IGEC-VI), 5-9 June, Eskisehir, Turkey.

2. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2011. Optimum operating conditions of the small-scale open and direct solar thermal Brayton cycle at various steady-state conditions. In: Proceedings of the 8th International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics (HEFAT2011), 11-13 July, Mauritius.

3. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2011. Maximum net power output of the recuperative open and direct solar thermal Brayton cycle. In: Proceedings of the 5th International Conference on Energy Sustainability (ASME, ES 2011), 7-10 August, Washington, USA.

4. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2012. Optimum small-scale open and direct solar thermal Brayton cycle for Pretoria, South Africa. In: Proceedings of the 1st Southern African Solar Energy Conference (SASEC 2012), 21-23 May, Stellenbosch, South Africa.

5. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2012. Optimum small-scale open and direct solar thermal Brayton cycle for Pretoria, South Africa. In: Proceedings of the 6th International Conference on Energy Sustainability (ASME, ES 2012-91135), 23-26 July, San Diego, California, USA.

6. Le Roux, W.G., Mwesigye, A., Bello-Ochende, T., Meyer, J.P., 2014. Tracker and collector for an experimental setup of a small-scale solar thermal Brayton cycle. In: Proceedings of the 2nd Southern African Solar Energy Conference (SASEC 2014), 27-29 January, Port Elizabeth, South Africa.

7. Le Roux, W.G., Bello-Ochende, T., Meyer, J.P., 2014. Optimisation of an open rectangular cavity receiver and recuperator used in a small-scale solar thermal Brayton cycle with thermal losses. In: Proceedings of the 10th International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics (HEFAT2014), 14-16 July 2014, Orlando, Florida, USA.

8. Le Roux, W.G., Meyer, J.P., Bello-Ochende, T., 2015. Experimental testing of a tubular cavity receiver for a small-scale solar thermal Brayton cycle (SASEC 2015), 11-13 May, Skukuza, Kruger National Park, South Africa.

79 Thank you

Questions?

[email protected]