EXPERIMENTAL INVESTIGATIONS OF HEAT … 082.pdf2nd International Seminar on ORC Power Systems,...

2nd International Seminar on ORC Power Systems, Rotterdam (Netherlands)

EXPERIMENTAL INVESTIGATIONS OF HEAT

TRANSFER CHARACTERISTCS AND

THERMAL STABILITY OF SILOXANES

Florian Heberle, Markus Preißinger, Theresa Weith and Dieter Brüggemann

Introduction Siloxanes as working fluids in ORC Power Systems

08.10.2013 EXPERIMENTAL INVESTIGATIONS OF HEAT

TRANSFER CHARACTERISTICS AND THERMAL

STABILITY OF SILOXANES - F. Heberle et al. Page 2

• Siloxanes are potential working fluids for ORC power systems.

• Advantages: long-term experiences, low toxicity and GWP = 0.

• Mainly used as ORC working fluids for high-temperature heat sources

like biomass-fired power plants or waste heat recovery units.

Experimental investigation

Thermal stability Heat transfer coefficient

• Comparison to correlations

• Economic evaluation (pure

fluids and mixtures)

• Maximum process

temperatures

• Decomposition products

Introduction Investigated working fluids




• Hexamethyldisiloxane (MM); n = 0

• Octamethyltrisiloxane (MDM); n = 1

• Decamethyltetrasiloxane (MD2M); n = 2

Fluid properties: T,s-diagram:

Structural

formula

Tcrit

(°C)

pcrit

(bar)

MM C6H18OSi2 245.6 19.4

MDM C8H24O2Si3 290.4 14.2

MD2M C10H30O3Si4 326.3 12.3 -1.0 -0.5 0.0 0.5 1.00

100

200

300

tem

pe

ratu

re (

°C)

entropy (kJ/(kgK))

MM

MDM

MD2M

Heat transfer characteristics Experimental setup




• pmax = 25 bar

• Tmax = 260 °C

Test conditions:

• 𝑞 = 8 – 18 kW/m2

• 𝐺 = 50 – 400 kg/(m2s)

• Electrical heated

steel pipe (DC power)

• Length: 5 m

10 m

m

12 m

m

3 thermocouples

electrical

insulation

DC power supply

test section

inlet

test section

outlet

P/T

1.1

P/T

1.2

Heat transfer characteristics Evaporation – Test section




Data reduction:

TTC

j = 0 j = 1 j = 2 TW,i

Tsat

ℎ𝑗 =𝑞

𝑇W,𝑖 − 𝑇sat(𝑝)

𝑇𝑊,𝑖 = 𝑇 𝑊,𝑜 +𝑞 i4𝜆

∙ (𝑟o2 − 𝑟i

2) +𝑞 i2𝜆

∙ 𝑙𝑛𝑟i𝑟𝑜

∙ 𝑟o2

𝑇 𝑊,𝑜 =𝑇𝑇𝐶,𝑡𝑜𝑝 + 2 ∙ 𝑇𝑇𝐶,𝑚𝑖𝑑𝑑𝑙𝑒 + 𝑇𝑇𝐶,𝑏𝑜𝑡𝑡𝑜𝑚

4

j = 10

Results Variation of mass flux density – MM




• h increases with

increasing mass

flux density

• h decreases with

increasing

vapour quality

0.0 0.2 0.4 0.6 0.8 1.0

0

2

4

6

8

10

12

14

he

at

tra

nsfe

r co

eff

icie

nt

(kW

/m2K

)

vapour quality (-)

G (kg/(m2s))

50

100

200

300

400

MM

p = 9 bar;

q = 8 kW/m2

Results Variation of heat flux density – MM




• No significant

influence of heat

flux density

• h decreases

with increasing

vapour quality

0.0 0.2 0.4 0.6 0.8 1.0

0

1

2

3

4

5

6

heat

transfe

r coeff

icie

nt

(kW

/m2K

)

vapour quality (-)

q ( kW/m2)

8

10

12

14

15.9

17

MM

G = 200 kg/(m2s);

Tsat

= 220 °C

Results Variation of examined working fluid – statistical and systematic uncertainties




• Different behaviour

of MM and MDM

depending on

vapour quality

• Statistical

uncertainties

(5 repetitions)

• Systematic

uncertainties

(ΔA/A; ΔP/P,

ΔTW,o/TW,o,

Δpsat/psat)

0.0 0.2 0.4 0.6 0.8 1.0

0

1

2

3

4

5

6

he

at

tra

nsfe

r co

eff

icie

nt

(kW

/m2K

)

vapour quality (-)

MM

MDM

Tsat

= 198 °C; G = 200 kg/(m2s); q = 15.9 kW/m

2

Results




Comparison to correlations

0.0 0.2 0.4 0.6 0.8 1.0

0

1

2

3

4

5

6

heat

transfe

r coeff

icie

nt

(kW

/m2K

)

vapour quality (-)

Experimental Data

Kandlikar (1998)

Saitoh et al. (2007)

MM; G = 200 kg/(m2s); T

sat = 198 °C; q = 15.8 kW/m

2

Results Comparison to correlations




0.0 0.2 0.4 0.6 0.8 1.0

0

1

2

3

4

5

6

heat

transfe

r coeff

icie

nt

(kW

/m2K

)

vapour quality (-)

Experimental Data

Kandlikar (1998)


MM; G = 200 kg/(m2s); T

sat = 198 °C; q = 15.8 kW/m

2

Results





0.0 0.2 0.4 0.6 0.8 1.0

0

1

2

3

4

5

6

heat

transfe

r coeff

icie

nt

(kW

/m2K

)

vapour quality (-)

Experimental Data

Kandlikar (1998)


MDM; G = 200 kg/(m2s); T

sat = 198 °C; q = 15.8 kW/m

2

Results




Comparison to correlations – working fluid: MM

• All measured

local h

• Mean relative

deviation (Kandlikar)

25.1 %

• Saitoh et al.

49.0 %

• Mean relative

deviation (MDM –

Kandlikar)

40.9 %

0 1 2 3 4 5 6 70

1

2

3

4

5

6

7

- 40%

experim

enta

l heat

transfe

r coeff

icie

nt

(kW

/m2K

)

calculated heat transfer coefficient (kW/m2K)

+ 40%

Kandlikar

Heat transfer measurements Main results




• Heat transfer coefficients are measured for process temperatures up to 250 °C.

• Empirical model of Kandlikar shows a good agreement to the experimental data.

Experimental investigation

Thermal stability Heat transfer coefficient

• Comparison to correlations

• Economic evaluation (pure

fluids and mixtures)

• Maximum process

temperatures

• Decomposition products

Thermal stability Experimental setup




• pmax = 30 bar

• Tmax = 500 °C

Test conditions:

• 𝑡 = 72 h

• 𝑇 = 240 – 420 °C

• Electrical heated by

heating wire

• Analysed by

gas chromatography/

mass spectroscopy

Thermocouple Pressure

device

Heating

wire

Results




Liquid phase, 360 °C, 144 h

• Fluid: Wacker® AK 0.65

• Purity: > 97 mass-%

• Formation of higher

chained siloxanes in

accordance to

Dvornic, Gelest, Inc.

time (min)

counte

r (-

) counte

r (-

) before

after

D5

MM

MDM

MD2M

MD4M

MD3M

MM

Results




Gas phase, 72 h

• Averaged molar

concentration before

tests: 99.4 mol-%

• Formation of

methane and ethane

in accordance to

Manders and Bellama,

Journal of Polymer

Science, 1985

Conclusions and Future work

• Heat transfer and thermal stability measurements were carried out

for selected siloxanes.

• The correlation of Kandlikar shows the best agreement to experimental data.

• No significant amount of decomposition products for heat transfer test conditions.

• Heat transfer characteristics of the mixture MM/MDM and MM/MDM/MD2M.

• Investigation of enhanced tubes and alternative working fluids.

• Long-term and dynamic tests concerning thermal stability.




Acknowledgements

The authors gratefully acknowledge financial support from




Free provision of Wacker® AK 0.65

“Fluid mixtures for efficiency increase of

Organic Rankine Cycles in selected

applications” (Grant no. 1713/12-1 and -2)

Partial financing of the

thermal stability test rig

www.zet.uni-bayreuth.de

Thank you

Florian Heberle, Markus Preißinger, Theresa Weith and Dieter Brüggemann

Heat transfer characteristics Evaporation – Test section




10

mm

12

mm

3 thermocouples TC

electrical

insulation

DC power supply

test section

inlet

test section

outlet

P/T

1.1

P/T

1.2

i = 1 – 10 '

'' '

ii

h hx

h h

1 1i

i i i

TF

Ph h h h

m 0 ( 0.5 )sath h T K

Data reduction:

TTC

subcooled

i = 0 i = 1 i = 2 TW,i

Ts

Results Variation of saturation pressure- MM




0.0 0.2 0.4 0.6 0.8 1.0

0

1000

2000

3000

4000

5000

6000

7000

hea

t tr

ansf

er c

oef

fici

ent

(W/m

2K

)

vapour quality (-)

psat

(bar)

5.8

7.75

9

13

MM

G = 200 kg/(sm2);

q = 15.8 kW/m2

Results Variation of examined working fluid




0.0 0.2 0.4 0.6 0.8 1.0

0

1

2

3

4

5

6hea

t tr

ansf

er c

oef

fici

ent

(kW

/m2K

)

vapour quality (-)

MM

MDM

MD2M

Tsat

= 220 °C;

G = 200 kg/(m2s);

q = 15.9 kW/m2

Results




Comparison to correlations – Model of Kandlikar

ℎ𝑡𝑐tp = ℎ𝑡𝑐tp,nbd ℎ𝑡𝑐tp,cbd

ℎ𝑡𝑐tp,nbd = 0.6683 ∙ 𝐶𝑜−0.2 ∙ 1 − 𝑥 0.8 ∙ ℎ𝑡𝑐LO + 1058 ∙ 𝐵𝑜0.7 ∙ 1 − 𝑥 0.8 ∙ 𝐹fl∙ ℎ𝑡𝑐LO

ℎ𝑡𝑐tp,cbd = 1.136 ∙ 𝐶𝑜−0.9 ∙ 1 − 𝑥 0.8 ∙ ℎ𝑡𝑐LO + 667.2 ∙ 𝐵𝑜0.7 ∙ 1 − 𝑥 0.8 ∙ 𝐹fl ∙ ℎ𝑡𝑐LO

𝐵𝑜 =𝑞

𝐺 ∙ ∆ℎ=

𝐴cs ∙ 𝑞

𝑚 ∙ ℎ′′ − ℎ′ 𝐶𝑜 =

𝜌g

𝜌l

0,51 − 𝑥

𝑥

0,8

ℎ𝑡𝑐LO =

𝜁2 𝑅𝑒LO − 1000 ∙ 𝑃𝑟𝑙

1,0 + 12,7 ∙𝜁2 𝑃𝑟l

23 − 1

∙𝜆l𝑑i

Results




Comparison to correlations – Model of Saitoh et al.

ℎ𝑡𝑐tp = 𝐹ℎ𝑡𝑐𝑙 + 𝑆ℎ𝑡𝑐𝑝𝑜𝑜𝑙

ℎ𝑡𝑐𝑙 = 0.223𝜆𝑙𝐷

∙𝐺 1 − 𝑥 𝐷

𝜂𝑙

0.8

∙𝑐𝑝,𝑙𝜂𝑙𝜆𝑙

0.8

𝐹 = 1 +1

𝑥

1.05

+ 1 +𝑊𝑒𝑔−0.4

S = 1/1 + 𝑅𝑒𝑡𝑝 ∙ 10−41.405

ℎ𝑡𝑐𝑝𝑜𝑜𝑙 = 207𝜆𝑙𝑑𝑏

∙𝑞 𝑑𝑏𝜆𝑙𝑇𝑙

0.745

∙𝜌𝑔

𝜌𝑙

0.581

∙ 𝑃𝑟𝑙0.533

Results




Comparison to correlations – working fluid: MM

• Measured

local htc

• Mean relative

deviation (Kandlikar)

25,1 %

• Saitoh et al.

49,0 %

0 1 2 3 4 5 6 70

1

2

3

4

5

6

7

- 40%

exp

erim

enta

l

hea

t tr

ansf

er c

oef

fici

ent

(k

W/m

2K

)

calculated experimental heat transfer coefficient (kW/m2K)

+ 40%

Saitoh et al.

Results Homogenous temperature profile




0 1 2 3 4 5

0

20

40

60

80

100

120

140

inner

wal

l te

mper

ature

(°C

)

length of test section (m)

t1

t2

t3

t4

Results




0 1 2 3 4 5

195

200

205

210

215

220

225

230

inn

er w

all

tem

per

atu

re (

°C)


MM top MM bottom

Results




0 1 2 3 4 5

195

200

205

210

215

220

225

230

inn

er w

all

tem

per

atu

re (

°C)


MDM top MDM bottom

Results Flow regimes - MM




0

50

100

150

200

250

300

350

400

450

500

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

G [

kg/m

²s]

x

x_m_IA

m_wavy

m_mist

m_strat

m_bubbly

Annular

Intermittent

Mist flow

Stratified wavy

Stratified

Results Flow regimes – MDM




0

50

100

150

200

250

300

350

400

450

500

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

G [

kg/m

²s]

x

x_m_IA

m_wavy

m_mist

m_strat

m_bubbly

Results Fluid properties




psat

(bar) pred

ρv

(kg/m3)

ρl / ρv

σ

(N/m)

MM 9.03 0.47 53.54 10.05 0.0154

MDM 2.90 0.21 20.82 29.44 0.0166

• Higher vapour density for MM lower vapour velocity at same mass flux.

• Nucleate dominates at low vapour qualities, caused by low surface tension and

liquid-to-vapour density ratio.

• Lower surface tension increase the probability of liquid entrainment in the vapour

core.

• Suppression of nucleate boiling is delayed by higher vapour density (lower velocity)

Thermal stability Temperature distribution




temperature (°C)

heig

ht

of

reacto

r(-)

Outline



STABILITY OF SILOXANES - F. Heberle et al.

Test procedure

Outline



STABILITY OF SILOXANES - F. Heberle et al.

Dynamic test rig

EXPERIMENTAL INVESTIGATIONS OF HEAT … 082.pdf2nd International Seminar on ORC Power Systems,...

Documents

Transcript of EXPERIMENTAL INVESTIGATIONS OF HEAT … 082.pdf2nd International Seminar on ORC Power Systems,...