carbon dioxide releases from a small scale pipeline - [PDF Document] (2024)

An experimental investigation on pressure response and phase transition of supercritical

carbon dioxide releases from a small-scale pipeline

Xingqing Yana, Hailong Zhua, Jianliang Yua *, Shaoyun Chenb, Haroun Mahgereftehc

a School of Chemical Machinery and Safety, Dalian University of Technology, Dalian, 116024,

China

b School of Chemical Engineering, Dalian University of Technology, Dalian, 116024, China

c Department of Chemical Engineering, University College London, London WC1E 7JE, UK

Abstract: The prediction of the pressure response and phase transition in the event of an

accidental carbon dioxide (CO2) release from a ruptured pipeline is of significant importance for

understanding the depressurization behaviour and hence the fracture behaviour. This article

presented a small-scale experimental investigation on the pressure response and phase transition

of supercritical CO2 release from a pressurized pipeline with a relief orifice. High-frequency

transducers and thermocouples were used to measure the evolution of CO2 pressures and

temperatures at different locations after release. The results indicated that pressures at different

locations decreased nearly synchronously after release. No vapour bubble and pressure rebound

generated in larger-scale release experiments were found in our small-scale release experiments.

The depressurization rate was greatly affected by the phase transition. During the release process,

the supercritical CO2 firstly turned into an unstable gas with a very great depressurization rate,

then changed into the gas–liquid phase with a lower depressurization rate, and finally changed

* Corresponding author: Tel.: +86-411-84986281; Fax: +86-411-84986281.

E-mail address: [emailprotected] (J. Yu)

into gaseous CO2. The larger the relief diameter was, the longer the gas–liquid phase state lasted.

Keywords: Supercritical CO2 release, pressure response, phase transition, small-scale pipeline

1 Introduction

One of the most difficult global environmental problems which human beings are now

facing is the increasing atmospheric greenhouse gases and the resulting global warming [1]. CO2

emitted from fossil fuel combustion is a major contributor to the greenhouse effect. This

situation will not be changed in the coming decades due to the actual energy situation of the

global energy structure [2]. Carbon capture and storage (CCS) technology intends to capture the

released CO2 at the emission sources, and transport the captured CO2 to storage locations to

mitigate the amount of CO2 released into the atmosphere [3-5].

The scale and safety requirements of CCS application determine that pipeline transportation

is the primary means of CO2 transportation, due to its high efficiency and good economy [6, 7]. It is

reported that a great deal of pipelines will need to be constructed in more densely populated

areas, where multiple anthropogenic sources exist [8]. Pipelines usually suffer from failure risks,

either puncture or full-bore rupture, caused by mechanical damage, corrosion, material defects,

or operational error [9, 10]. After failure, CO2 is released suddenly from the pipeline, causing

property loss and casualties owing to asphyxia.

CO2 released from pressurized pipelines is more complicated to deal with than other

substances, because the CO2 release process may include a combination of gaseous, liquid, and

solid state CO2 [11, 12]. During the CO2 depressurization process after accidental release from an

initially liquid or supercritical state to ambient conditions, the pressure and temperature drop

significantly owing to the expansion and subsequent high Joule–Thomson cooling effect, which

causes the complex phase transition of CO2 [13]. There have been several large research projects

focusing on CO2 release behaviour, including COOLTRANS [14], CO2PipeTrans [15], CO2PipeHaz [8],

CO2QUEST [16], COSHER [17], and so on. This research is very helpful for understanding the safety

issues related to CO2 transportation.

Numerous publications have examined the release behaviour of CO2 based on both

experimental studies and numerical research. We have given a detailed introduction in our

recently published works [18-22] and will therefore not repeat it in this article. Despite numerous

studies nowadays, there is not yet a clear understanding of the pressure response and phase

transition of the supercritical CO2 depressurization process after accidental release from a

pressurized pipeline. We performed an experimental study on the pressure response and phase

transition of supercritical CO2 during sudden release to improve the understanding of this process

using a large-scale pipeline with a length of 258 m and an inner diameter of 233 mm from the

CO2QUEST project, and found that the complex phenomena of pressure undershoot, rebound or

slowdown occurred near the critical region [19, 20]. Moreover, we discussed the phase transitions

of CO2 in the pipeline at different diameters of relief orifices during the release process [21,22].

Medium-scale CO2 release experiments were also performed to focus on the phases in the

releases when supercritical and saturation CO2 were released from a 5 m pipe connected to a 2

m3 spherical vessel. Five stages with different depressurization rates and phase states were

analyzed [23, 24]. However, these studies are far from sufficient.

With regard to the aforementioned problems, a study was proposed to focus on the

pressure response and phase transition of supercritical CO2 released from a small-scale

pressurized pipeline. High-frequency transducers were used to record the evolutions of fluid

pressures during release. Thermocouples were placed to measure the temperatures of fluid.

2 Experiments

2.1 Experimental set-up

To investigate the pressure response and phase transition of supercritical CO2 from a

pipeline, a small-scale experimental facility was established [18]. The schematic diagram and

photograph are shown in Figure 1. It consisted of a CO2 insulated Dewar vessel, a 10 L buffer tank,

a main pipe, a relief pipe, and a pneumatic valve driven by compressed air from a gas cylinder.

CO2 from the Dewar vessel was conditioned into the 10 L buffer tank, which was coated by a heat

band which kept the CO2 inside the buffer tank at the desired temperature, and glass wool

insulation which kept the whole leakage process inside the buffer tank under a near-adiabatic

condition.

A main pipe with an internal diameter of 25 mm and a length of 5 m was connected to the

bottom of the buffer tank by an elbow. The other end of the main pipe was connected to a

pneumatic valve. Several relief pipes with internal diameters of 25 mm and lengths of 1 m were

machined before the experiments. These pipes were all closed at one end and each had a circular

hole in the middle to simulate the leakage nozzle. When doing experiments, a selected relief pipe

with the desired circular hole diameter was assembled to the pneumatic valve. Both the main

pipe and relief pipe were coated by a heat band and glass wool insulation. The circular hole

diameters in this study were 2 mm, 4 mm and 6 mm.

The pneumatic ball valve was DQ641Y type used in low temperature conditions. It could be

used with the pressure less than 10 MPa and the temperature larger than -100 °C. Its duration

time during value opening was about 0.1s~1s, depending on the actuating pressure.

Three pressure transducers labelled P1, P2, and P3 and three armoured K-type

thermocouples labelled T1, T2, T3 were mounted at different locations along the main pipe to

record the pressures and temperatures of CO2 inside the main pipe, as shown in Figure 2. The

accuracy and frequency of the pressure transducers were ±0.25% and 100 kHz. The response

time and uncertainty of the thermocouples are 1 s and ± 1 °C. All sensors were calibrated before

use. The data acquisition system was accomplished by the NI acquisition module and LabVIEW

software [18-22].

2.2 Experiment procedure

The following steps are involved in each test: (1) Assembling the relief pipe with the desired hole

diameter; (2) Examining the integrity of the whole pipeline set-up; (3) Debugging the testing

instruments and data acquisition system; (4) Opening the atmospheric exhaust valve of the buffer

tank, filling gas phase CO2 from the Dewar vessel to the buffer tank for about 30 s, and then

closing the exhaust valve and ending the filling process (keeping a pressure of 0.5 MPa in the

buffer tank); (5) Filling liquid phase CO2 from the Dewar vessel to the buffer tank. During this step,

the evacuation of gas CO2 in the buffer tank and the filling of liquid CO2 to the buffer tank might

be performed alternately. Experience was needed to judge the proper amount of CO2 filled; (6)

After the filling process, opening the heating system to make the pressure and temperature of

CO2 in the buffer tank meet the required experimental conditions. Evacuation might be needed in

this step; (7) Opening the data acquisition system, setting up the image device; and activating the

pneumatic valve; (8) Clearing up the experimental field.

Special attention was needed to safety issues during each step. Each test was repeated

several times (usually three) to ensure repeatable results within the permitted error range. All

measurements were carried out under similar ambient temperature (about 15 °C) and humidity

(about 60%). The initial pressure of 9 MPa and the initial temperature of 40 °C were selected in

every experiment. Three relief diameters of the leakage nozzle were chosen: 2 mm, 4 mm, and 6

mm.

3 Results and discussions

3.1 Pressure and temperature evolutions in main pipe

Figure 3 shows the pressure evolutions at different locations inside the main pipe in the

supercritical CO2 release experiments with relief diameters of 2 mm (a), 4 mm (b), and 6 mm (c).

In every subgraph, the release began at time t1 when the pneumatic valve was activated. The

pressures inside the main pipe underwent a sharp drop due to the expansion and release of

supercritical CO2 after the pneumatic valve was opened. The precipitous declines of pressure

were slowed at time t2. Apparently stable and gentle pressure drops appeared after t2. After a

period of time, the depressurization rate changes again at time t3.

Based on the variation trend of the depressurization rate, three stages could be

distinguished during every release process. The first stage was from time t1 to time t2. The

durations were about 5.5717 s for 2 mm release diameter, 1.0365 s for 4 mm release diameter,

and 0.752 s for 6 mm release diameter. The second stage was from time t2 to time t3. The

durations were about 63.1628 s for 2 mm release diameter, 19.6248 s for 4 mm relief diameter,

and 9.3560 s for 6 mm relief diameter. The third stage was from t3 to the end time of the release.

The different stages with various depressurization rates were attributed to the phase states of

CO2 inside the main pipe, which will be discussed in detail in the next section.

Moreover, the pressure curves were nearly synchronous at the testing locations P1, P2, and

P3 in each experiment. Though the pressure of P3 was slightly greater than that of P2, which was

slightly greater than that of P1, the pressure gradient from P3 to P1 was small. The synchronous

changes of fluid pressures at different locations were similar to those tested in large-scale

experimental pipelines with a length of 258 m and an inner diameter of 233 mm. However, no

vapour bubbles and pressure rebound [19,20] were found in the small-scale experiments. Also, the

propagation of decompression waves recorded in the large-scale experiments was harder to

estimate in this study, probably because the experimental pipeline was too short.

Figure 4 shows the temperature evolutions at different locations inside the main pipe in the

supercritical CO2 release experiments with relief diameters of 2 mm (a), 4 mm (b), and 6 mm (c).

Temperatures decreased just after the pneumatic valve was activated, due to the Joule–Thomson

effect of the expansion process. Locations near the release orifice showed lower temperatures

during the temperature decreasing process. But in these experiments, the temperature gradients

were not very large. The largest temperature drops reached at location T1 were about 38 °C for 2

mm relief diameter, 52 °C for 4 mm relief diameter, and 58 °C for 6 mm relief diameter. Larger

relief diameters brought bigger temperature drops.

3.2 Phase transformation during release

Combining the tested pressure data with the temperature data at different times, we

obtained the phase changes of CO2 during the release process, as shown in Figure 5. After the

pneumatic valve acted, the inventory properties inside the experimental main pipes (P1, T1; P2, T2;

P3, T3) passed through the supercritical region and into the gaseous region of the phase diagram,

due to the rapid pressure drop. Apparently, the supercritical state CO2 was unlikely to change to

dense phase state because the pressure drop was more dominant than the temperature drop.

Then the curves of all the measuring points tend to be close to the saturation line, which

suggested that the unstable gaseous CO2 quickly transformed into the gas–liquid phase. For a

large part of the release, the measured pressures and temperatures of all locations followed

approximately the saturation line. After a period of time along the saturation line, the pressure–

temperature developments of all the testing points along the pipeline deviated from the

saturation line, indicating that the gas–liquid phase CO2 transformed into gaseous CO2. The phase

transition process was similar to those obtained in the large-scale experimental pipeline with a

small release orifice [19] and medium-scale CO2 releases with full bore rupture conditions [23].

The reason for the different depressurization rates in the different stages represented in

Figure 3 can be explained as below. As shown in Figure 5, during the first stage from t1 to t2, the

pressure drop was caused by the supercritical CO2 release with a large depressurization rate.

Then, during the second stage from t2 to t3, the supercritical CO2 changed into gas–liquid phase

CO2, which slowed the depressurization rate of CO2, due to the smaller density and larger

compressibility of gas–liquid phase CO2 than supercritical CO2. Finally, during the third stage from

t3 to the release end, the pressure drop was attributed to the gas phase CO2 release, during

which the depressurization rate was different compared to the process of gas–liquid phase CO2

release. In Figure 5, the valve acting time t1, the first phase transformation time t2, and the

second phase transformation time t3 were plotted. The corresponding pressures of t1, t2, and t3 in

Figure 5 were similar to those of t1, t2, and t3 in Figure 3, demonstrating that the changes of the

depressurization rate were related to the phase transformation.

3.3 Jet plume outside the relief orifice

High pressure CO2 decompresses rapidly outside the leakage orifice, and experiences an

expansion process, leading to a typical under-expanded plume in the dispersion process, which

can be seen clearly as a white plume [25]. Figure 6 and Figure 7 show the appearance of jet

dispersion at different times after release in the experiment with the relief diameters of 2 mm

(Figure 7) and 4 mm (Figure 7). Due to the violent temperature drop caused by the J–T effect

during the release process, the extremely low temperature (usually lower than -78 °C for

supercritical CO2 release) led to a phase change of CO2 from other phase states to solid phase. As

shown in Figure 6 and Figure 7, a compact core area inside the jet plume near the release orifice

was formed, where a large fraction of solid CO2 particles was generated. At 0 s after release

(Figure 6a and Figure 7a), supercritical CO2 in the pipe was released from the orifice. Then due to

the transition from supercritical CO2 to unstable gaseous CO2 in the pipe, the dimensions of the

jet plume decreased (Figure 6b, c, d and Figure 7b). Some time later, gaseous CO2 in the pipe was

transformed to gas–liquid phase, and the dimensions of the jet plume became larger and larger

(Figure 6e, f and Figure 7c, d). After the gas–liquid phase CO2 changed into gaseous CO2 in the

pipe, the dimensions of the jet plume gradually became smaller and invisible (Figure 6e, f and

Figure 7 e).

3.4 Discussions

Recently we performed both larger-scale and small-scale supercritical CO2 release

experiments to study the pressure response and phase transition of CO2 inside the experimental

pipe [19-22]. Also, the results of medium-scale supercritical CO2 release experiments can be found

in the literature [23, 24]. There were both similarities and differences among the different scale

release experiments.

First, the pressure responses during pipeline depressurization were different. The pressure

undershoot, rebound or slowdown which occurred near the critical region in the large-scale

experiments of supercritical CO2 release [19,20] were not found in the medium-scale [23] and

small-scale supercritical CO2 release experiments. There may be a variety of possibilities for this

difference, but the following two were the most likely reasons.

(1) In larger-scale CO2 release experiments, the initiation of release was controlled by two

blasting discs [19-22], which could be fully open in several millisecond. However, in small-scale CO2

release experiments, the initiation of release was activated by a pneumatic ball valve, which had

a much lower opening speed (nearly one second for fully opening). The faster the release orifice

opened, the more severe the depressurization wave propagated and rebounded. Hence the

pressure undershoot, rebound or slowdown would occur more easily in larger-scale experiments.

(2) In large-scale CO2 release experiments, the pipe internal diameter of 243 mm was large

enough that temperature gradient existed on the same cross section at the release onset [19-20].

However, in small-scale CO2 release experiments, the internal diameter of the vent pipes was

only 25 mm, and was too small to have temperature gradient on the same cross section. Hence

the intense heat and mass transfers and the resulting pressure undershoot and rebound

phenomenon occurred in the large-scale release experiments at the release onset were not seen

in the small-scale experiments.

Second, the phase transitions of supercritical CO2 during pipeline depressurization were

similar [19, 23]. The initial supercritical CO2 transformed into gaseous CO2 (normally this stage was

transient and unstable), followed immediately by the transformation of CO2 into gas–liquid phase.

After a period of time, the pressures and temperatures successively deviated from the saturation

line and the gas–liquid CO2 transformed into gas phase.

4 Conclusions

This article has presented the results of a small-scale experimental investigation on the

pressure response and phase transition of supercritical CO2 release from a pressurized pipeline

with a relief orifice. Some conclusions are demonstrated as follows:

(1) The pressure undershoot, rebound or slowdown which occurred near the critical region

in the large-scale experiments of supercritical CO2 release were not found in the small-scale

supercritical CO2 depressurization process.

(2) Locations near the release orifice reached lower temperatures during the temperature

decreasing process. Larger relief diameters brought bigger temperature drops.

(3) Phase transitions occurred during the supercritical CO2 release process. The supercritical

CO2 firstly turned into an unstable gas with a very great depressurization rate, then changed into

the gas–liquid phase, and finally changed into gaseous CO2.

(4) The larger the relief diameter was, the longer the gas–liquid phase lasted during the CO2

release process.

(5) Both similarities and differences in the pressure responses and the phase transitions

were found among the different scale release experiments.

Acknowledgement

The authors would like to acknowledge the funding received from the European Union

Seventh Framework Programmes FP7- ENERGY-2009-1 under grant agreement number 241346

and FP7-ENERGY-2012-1STAGE under Grant agreement 309102.

References

[1] Putman WM, Ott L, Darmenov A, DaSilva A. A global perspective of atmospheric carbon

dioxide concentrations. Parallel Comput 2016; 55:2–8.

[2] Myers SS, Wessells KR, Kloog I, Zanobetti A, Schwartz J. Effect of increased concentrations of

atmospheric carbon dioxide on the global threat of zinc deficiency: a modelling study.

Lancet Glob Health 2015; 3: 39-45

[3] Haszeldine RS. Carbon capture and storage: How green can black be?. Science 2009; 325:

1647–1652.

[4] Kanniche M, Gros-Bonnicard R, Jaud P, Valle-Marcos J, Amann J, Bouallou C. Pre-combustion,

post combustion and oxy-combustion in thermal power plant for CO2 capture. Appl Therm

Eng 2010; 30: 53–62.

[5] Duan HB, Fan Y, Zhu L. What’s the most cost-effective policy of CO2 targeted reduction: an

application of aggregated economic technological model with CCS?. Appl Energ 2013; 112:

866–75.

[6] Munkejord ST, Hammer M, Løvseth SW. Intergovernmental panel on climate change, carbon

capture & storage. Appl Energ 2016; 169: 499–523.

[7] Knoope MMJ, Ramirez A, Faaij APC. A state-of the-art review of techno-economic models

predicting the costs of CO2 pipeline transport. Int J Greenh Gas Con 2013; 16: 241–270

[8] Woolley RM, Fairweather M, Wareing CJ, Falle SAG, Mahgerefteh H, Martynov S, Brown S,

Narasimhamurthy VD, Storvik IE, Saelen L, Skjold T, Economou IG, Tsangaris DM,

Boulougouris GC, Diamantonis N, Custo L, Wardman M, Gant SE, Wilday J, Zhang YC, Chen

SY, Proust C, Hebrard J, Jamois D. CO2PipeHaz: Quantitative hazard assessment for next

generation CO2 pipelines. Energy Procedia 2014; 63: 2510–2529.

[9] Koornneef J, Spruijt M, Molag M, Ramirez A, Turkenburg W, Faaij A. Quantitative risk

assessment of CO2 transport by pipelines – A review of uncertainties and their impacts. J

Hazard Mater 2010; 177: 12–27.

[10] Duncan IJ, Wang H. Estimating the likelihood of pipeline failure in CO2 transmission pipelines:

New insights on risks of carbon capture and storage. Int J Greenh Gas Con 2014; 21: 49–60

[11] Elshahomi A, Lu C, Michal G, Liu X, Godbole A, Venton P. Decompression wave speed in CO2

mixtures: CFD modelling with the GERG-2008 equation of state. Appl Energ 2015; 140: 20–

32.

[12] Hill TA, Fackrell JE, Dubal MR, Stiff SM. Understanding the consequences of CO2 leakage

downstream of the capture plant. Energy Procedia 2011; 4: 2230–2237.

[13] Mazzoldi A, Hill T, Colls JJ, Assessing the risk for CO2 transportation within CCS projects, CFD

modelling. Int J Greenh Gas Con 2011; 5: 816–825.

[14] Copper R. National grid’s COOLTRANS research programme. J Pipel Eng 2012; 11: 155–172.

[15] Brown J, Holt H, Helle K. Large scale CO2 releases for dispersion model and safety study

validation. Energy Procedia 2014; 63: 2542–2546.

[16] Brown S, Martynov S, Mahgerefteh H, Fairweather M, Woolley RM, Wareing CJ, Falle SAG,

Rutters H, Niemi A, Zhang YC, Chen SY, Besnebat J, Shah N, Dowell NM, Proust C, Farret R,

Economou IG, Tsangaris DM, Boulougouris GC, Wittenberghe JV. CO2QUEST:

Techno-economic assessment of CO2 quality effects on its storage and transport. Energy

Procedia 2014; 63: 2622–2629.

[17] Ahmad M, Lowesmith B, Koeijer GD, Nilsen S, Tonda H, Sponelli C, Cooper R, Clausen S,

Mendes R, Florisson O. COSHER joint industry project: Large scale pipeline rupture tests to

study CO2 release and dispersion. Int J Greenh Gas Con 2015; 37: 340–353.

[18] Yan X, Guo X, Liu Z, Yu J. Release and dispersion behavior of carbon dioxide released from a

small-scale underground pipeline, J Loss Prevent Proc 2016; 43: 165-173

[19] Guo X, Yan X, Yu J, Zhang Y, Chen S, Mahgerefteh H, Martynov S, Collard A, Proust C. Pressure

responses and phase transitions during the release of high pressure CO2 from a large-scale

pipeline. Energy 2017; 118: 1066-1078.

[20] Guo X, Yan X, Yu J, Zhang Y, Chen S, Mahgerefteh H, Martynov S, Collard A, Proust C. Pressure

response and phase transition in supercritical CO2 releases from a large-scale pipeline. Appl

Energ 2016; 178: 189-197.

[21] Guo X, Yan X, Yu J, Zhang Y, Chen S, Mahgerefteh H, Martynov S, Collard A, Proust C.

Under-expanded jets and dispersion in high pressure CO2 releases from an industrial scale

pipeline. Energy 2017; 119: 53-66.

[22] Guo X, Yan X, Yu J, Zhang Y, Chen S, Mahgerefteh H, Martynov S, Collard A, Proust C.

Under-expanded jets and dispersion in supercritical CO2 releases from a large-scale pipeline,

Appl Energ 2016; 183: 1279-1291.

[23] Hebrard J, Jamois D, Proust C, Spruijt M, Hulsbosch-Dam CEC, Molag M, Messina E. Medium

scale CO2 releases. Energy Procedia 2016; 86: 479-488

[24] Jamois D, Proust C, Hebrard J. Hardware and instrumentation to investigate massive releases

of dense phase CO2. Can J Chem Eng 2015; 93: 234-240.

[25] Li K, Zhou X, Tu R, Xie Q, Yi J, Jiang X. An experimental investigation of supercritical CO2

accidental release from a pressurized pipeline. J Supercrit Fluid 2016; 107: 298-306

Figure 1 Schematic diagram of the supercritical phase CO2 release experimental set-up

Pneumatic value

P

Low temperature hose

10 L buffer tank

Levergauge

Main pipe(I. D. =25 mm, Length=5m)

Gas outlet Gas

inlet

Liquid outlet Liquid

inlet

CO2 insulatedDewar flask

Relief pipe (I. D. =25 mm, Length=1m)

Air cylinder

Leakage nozzleTest section

Figure 2 Schematic diagram of the leakage nozzle and testing locations on relief pipe and main

pipe

P1

Buffer tank

5 m

P2P3

T3 T2 T1 0.5 m

2 m 1.5 m

Leakage nozzle

Pneumatic valveMain pipe

Relief pipe1 m

Figure 3 Pressure evolutions of the supercritical CO2 release experiments with different relief

diameters ( initial pressure of 9 MPa, initial temperature of 40 °C).

0 20 40 60 80 1004

5

6

7

8

9

10

t3(78.2916s): Dividing point

of depressurization rate

Pre

ssu

re (

MP

a)

Time (s)

P1

P2

P3

(a) de=2mm

t1(15.1288s): Valve acts

t2(20.7005s): Dividing point

of depressurization rate

0 10 20 30 40 500

2

4

6

8

10

Pre

ssu

re (

MP

a)

Time (s)

P1

P2

P3

t1(1.9675s): Valve acts

t2(3.004s): Dividing point of

depressurization rate

(b) de=4 mm

t3(22.6288s): Dividing point

of depressurization rate

0 5 10 15 20 25 300

2

4

6

8

10

t3(18.9527s): Dividing point

of depressurization rate

Pre

ssu

re (

MP

a)

Time

P1

P2

P3

t1(8.8447s): Valve acts

t2(9.5967s): Dividing

point of

depressurization rate

(c) de=6 mm

Figure 4 Pressure and temperature evolutions of the supercritical CO2 release experiments with

different relief diameters (initial pressure of 9 MPa, initial temperature of 40 °C).

0 25 50 75 100 125 150 175 2000

10

20

30

40

50

T1

T2

T3

Tem

per

atu

re (

°C)

Time (s)

(a) de=2 mm

0 10 20 30 40 50 60 70 80-20

-10

10

20

30

40

50

Tem

per

atu

re (

°C)

Time (s)

T1

T2

T3

ΔTmax=52°C

(b) de=4mm

0 10 20 30 40 50 60-20

-10

10

20

30

40

50

T1

T2

T3

Tem

per

atu

re (

°C)

Time (s)

ΔTmax=58°C

(c) de=6mm

Figure 5 Pressures-temperature development of the supercritical CO2 release experiments (initial

pressure of 9 MPa, initial temperature of 40 °C).

2

4

6

8

10

-20 -10 0 10 20 30 40 50

t3: Phase transformation

point

Temperature (°C)

Pre

ssu

re (

MP

a)

T1, P1

T2, P2

T3, P3

t1: Valve acts

t2: Phase transformation

point

(a) de=2mm

2

4

6

8

10

-20 -10 0 10 20 30 40 50

t3: phase transformation

point

Temperature (°C)

Pre

ssu

re (

MP

a)

T1, P1

T2, P2

T3, P3

Saturation line

t1: Valve acts

t2: phase transformation

point

(b) de=4mm

-20 -10 0 10 20 30 40 500

2

4

6

8

10

t3: Phase

tranformation

point

Pre

ssu

re (

MP

a)

Temperature (°C)

T1, P1

T2, P2

T3, P3

t1: Valve acts

t2: Phase tranformation

point

Saturation line

(c) de=6mm

Figure 6 Jet flow phenomena (initial pressure of 9 MPa, initial temperature of 40 °C, and leakage

hole diameter of 2 mm)

Time after

release (

a) 0s

(

b) 1s

(

c) 5s

(d

) 12s

(e

) 25s

(e

) 51s

(e

) 73s

(e)

125s

Figure 7 Jet flow phenomena (initial pressure of 9 MPa, initial temperature of 40 °C, and leakage

hole diameter of 4 mm).

(a) 0s after

release

(b) 2s after

release

(c) 4s after

release

(d) 22s after

release

s

(e) 28s after

release