Two-phase flow through safety/relief valves, nozzles, orifices/VASIB facility


Activity description
Experimental facility - (Facility Performance )
Applications
Publications

Activity responsible: Dr. Gino BOCCARDI
Address: ENEA C.R. Casaccia, Energy Department, Institute of Thermal Fluid Dynamics, Via Anguillarese 301 (S.P. 092), 00060 S. M. di Galeria RM, Italy
Phone: +39 06 3048.3664 Fax: +39 06 3048.3026

Email:boccardi@casaccia.enea.it


Activity description

VASIB facility, designed to test two-phase flow through safety and relief valves, has a maximum operating pressure of 2.0 MPa, a maximum flow rate of 1500 kg/h, and an available electric power of 150 kW for the heating of the process fluid (water); vapour quality may range from 0 to 20 %.

Unlikely from most of available experimental facilities to study the two-phase flow through safety systems, the VASIB facility has been specifically designed to work as a closed loop, with the main aim of getting information on the effect of the back pressure on the mixture flow.

The facility can allow the testing of the two-phase flow through safety devices, with a back pressure different from the atmospheric value, using either actual safety valves (ø max = 10 mm) or reference geometries (convergent, divergent, or straight nozzles, etc). Besides the thermal-hydraulic characteristics of the VASIB facility allow to keep the test conditions in the loop without any time restriction, at low operating costs, thanks to the recovery heat exchanger.

To get these results, the facility can:

  • guarantee the head required from the specific test, independent of the fixed flow rate
  • have a sufficient thermal power available to reach and maintain the required operating conditions;
  • ary and control the test pressure and back pressure, acting on the pressurizer and on the opening of a regulating valve downstream the test section.

Top


Experimental facility

VASIB facility

Click here to see the complete scheme of the facility.

Top

Facility performance

c c

Main Loop

Cooling Tower Loop

Fluid

c

water

water

Mass flow rate (min/max)

kg/s

0.042 / 0.416

2.77

Vapour quality

%

0 / 20

-

Temperature (min/max)

°C

10 / 210

5 / 30

Pressure (min/max)

bar

2 / 20

2 / 4

Heater electric power (max)

kW

150

-

Top


Applications

1 st Test Campaign
2 nd Test Campaign
3 rd Test Campaign
4 th Test Campaign
5 th Test Campaign

Results obtained so far are reported below in.

Top


1 st Campaign

The main aim of the first test campaign has been the measurement of the two-phase mass flow rate and the correlated parameters, in order to achieve a first assessment of design correlations available for the sizing of the safety values under two-phase flow conditions.

An actual safety valve has been used as a test section ; it has an orifice diameter of 10 mm, and is instrumented with pressure transducers in the nozzle and in the valve body.

The lift of the valve plug may be set and imposed manually, in order to ascertain its influence on the discharge conditions. In the table here under the operating conditions tested are reported

c

SdP

Tin min/max (°C)

120/185

Tout min/max (°C)

119/180

Pin min/max (kPa)

200/1080

Pout min/max (kPa)

200/1000

Titin min/max (-)

0/15

Tit out min/max (-)

0/16

Port min/max (kg/h)

170/1500

Pot. min/max (kW)

20/145

1a Test Campaign: test conditions

Top


2 nd Campaign

In the second test campaign the steam-water data obtained with the VASIB facility have been compared with available data from the literature, most of which obtained using air-water mixtures.

Using air, as the gas phase, is an easy and effective way to simulate steam in a liquid flow, but has some drawbacks, in particular when the back pressure is higher than the atmospheric pressure (main feature of the VASIB facility). Indeed, as the air density is not so dependent on the pressure, the steam density shows large variations even form small changes in the system pressure beside, the expansion of the mixture implies also large variations in the quality. In simple words, the continuous expansion of the steam-water two-phase flow implies also variations in velocity, pressure drop and quality which, in turn, have a feedback on the expansion. Consequently, the evolution of the phenomenon, depending on all these inter-related factors, turns out to be very complex and difficult to be represented by an-water modeling.

In the present tests, we used a test section having a Venturi shape as it is among the most studied geometries for such a problem. In this way, the obtained data can be easily compared with similar data obtained using an air-water mixture.
The experimental data are also compared with the results of a calculation method based on the homogeneous equilibrium model. The results show that the model is quite reliable, with some uncertainties caused by the test procedures (the mass flow rate is imposed); indeed, this condition hampers the checking of the critical flow arising, in spite of the information derived from the test section pressure profile .

c

1st series

2nd series

3rd series

4th series

Flow rate kg/h

270

350

500

560

Pin min/max bar

5.45 / 6.58

4.36 / 7.48

5.7 / 11.65

7.15 / 8.3

Pout min/max bar

4

2.3 / 4.3

1.95 / 3.3

2.45 / 3

Xin min/max %

6 / 9.5

1 / 7

1 / 6

1 / 2

Xout min/max %

8 / 13

4.5 / 11

6 / 18

8.5 / 9.5

2nd Test Campaign: test conditions


3 rd Campaign

The tests are being carried out on a commercial PSV; the original øor=10mm has been reduced to a ø or=5mm, ( test section )  in order to obtain a greater quantity of experimental data. The lift of the valve plug is set at the value reported by constructor.

These data are compared, with various inlet parameters, with the predictions obtained using two method, one based on the Homogeneous Non Equilibrium Model hypothesis (HNE) and the other one on Homogeneous Equilibrium Model hypothesis (HEM). The comparison between the model predictions allows a qualitative evaluation on their performance in sizing of pressure safety valve in two-phase conditions.
In the table here under the operating conditions tested are reported.


1st series

2nd series

3rd series

4th series

5th series

Number of  tests
41 37
32
14
9

Flow rate  kg/h

280

350

500

750

1000

Pin min/max bar

4.1/17.4
4.9/17.33
4.9/17.1
4.2/14.2
6.1/13.6

Pout min/max bar

3/17.4
4/17.3
4/17
3.9/14
5.7/13.2

xin min/max %

5 K subcool./10
5 K subcool./9.9
5 K subcool./9.8
5 K subcool./0.9
5 K subcool./0.4

xout min/max %

5 K subcool./10.5
5 K subcool./11.9
5 K subcool./12.5
5 K subcool./3.2
5 K subcool./1.9

4th Campaign

In the previous campaign the geometry modification of PSV has reduced the number of possible tests and has not allowed a comparison with a real flow through a PSV. For these reasons a short campaign with the original orifice has be carried out; the results has confirmed the 3rd campaign evaluations.

In the table here under the operating conditions tested are reported.


1st series

2nd series

Number of  tests
15
21

Pin                   bar

5
7.5
Flow-rate         kg/h
350,500,750,1000
350,500,750,1000

Pout min/max    bar

2.2/4.9
3.1/7.4

xin min/max       %

5 K subcool./5
5 K subcool./10

xout min/max     %

5 K subcool./8.5
5 K subcool./13.7

5th Campaign

The lack of a reference standard represents a serious limit for PSV industrial applications in two-phase flow; in order to overcome it, different agencies are looking for a fairly simple correlation that considers all the two-phase flow aspects. Recently, correlations developed from the Homogeneous Equilibrium Model (HEM) have been considered the most interesting. For example, API (American Petroleum Institute) has edited in 2000 the API RP-520 "Sizing, Selection and Installation of Pressure Relieving Devices in Refineries” where a method for PSV sizing in two-phase flow, based on the HEM, is suggested.

In this campaign the tests are being carried out on a commercial PSV, øor=10mm, locked in the open position corresponding to the supplied discharge coefficients. The valve geometry has not been modified for obtaining the real working conditions. The experimental data have been compared to a lot of models, based on Homogeneous Non Equilibrium Model hypothesis (HNE) or HEM hypothesis, including the method suggested by API .
The mass flow-rate is calculated multiplying the model prediction for a theoretical nozzle by the discharge coefficient k s. Its importance is evident for computing the real mass flow-rate value; PSV manufactures supply and guarantee the coefficient for liquid (kl ) and gas (k g) while in two-phase conditions no information is directly available about ks because its value depending on the flow conditions.
In calculation a simple correlation that considers the inlet conditions and kl and kg, for discharge coefficient computing is proposed and an evaluation of its influence on mass flow-rate prediction has been done.

The test matrix proposed in the experimental campaign aims to gradually reach (or to approach as much as possible, based on the performance of the VASIB) conditions that limit the mass flow-rate through a safety device, which are low values of available pressure drop compared to flow rate and quality desired or critical flow conditions. The phenomenon variables are mass flow-rate, inlet pressure, inlet quality, outlet quality and back-pressure. In this experimental investigation, the inlet pressure, the mass flow rate and the inlet quality are kept constant: the back-pressure is measured while the outlet quality, as all the other quality values, is calculated by enthalpy balance.

To avoid using data affected by instrumentation errors, the tests with a pressure loss  through the valve of less than 20 kPa have not been considered as this value is too close to the pressure measurement uncertainty zone. This screening involved a reduction to 132 in the tests evaluated. The tests carried out and considered for each inlet pressures are shown in the following  table.
 


1st series

2nd series

3rd series

4th series

5th series

6th series

Nunber of  tests
15
16
22
29
25
25

Pin                    bar

5
7.5
10
12.5
15
17.5
Flow-rate         kg/h
350,500,750,
1000,1250,1500
350,500,750,
1000,1250,1500
350,500,750,
1000,1250,1500
350,500,750,
1000,1250,1500
350,500,750,
1000,1250,1500
350,500,750,
1000,1250,1500

Pout min/max    bar

3.7/4.8
6/7.3
7.4/9.8
6.8/12.3
11.5/14.8
13.8/17.3

xin min/max       %

5 K subcool./10
5 K subcool./ 10
5 K subcool./10
5 K subcool./10
5 K subcool./10
5 K subcool./10

xout min/max     %

5 K subcool./10.7
5 K subcool./11.4
5 K subcool./12.6
5 K subcool./13.9
5 K subcool./11.1
5 K subcool./11.6

Examples of data evaluation with the method suggested by API (HEM) and with a model (Henry and Fauske) developed from HNE  hypothesis

Top


Publications

G.P. CELATA and G. GUIDI, Problems About the Sizing of Two-Phase Flow Safety Valves
Heat & Technology, Vol. 14, n. 1, pp. 67-96, 1996
Boccardi G., Celata G.P., Iorizzo A. "Risultati dell’attività sperimentale sulle valvole di sicurezza in bifase ottenuti con l’impianto VASIB" doc. CNR, Dicembre 1997
Boccardi G., Celata G.P., Iorizzo A. - "Risultati dell’attività sperimentale sulle valvole di sicurezza in bifase ottenuti con l’impianto VASIB - Seconda parte" doc. CNR, Dicembre 1998
Alimonti C., Boccardi G., Celata G.P. -"Water Two-Phase Flow Through a Convergent-Divergent Nozzle With Variable Backpressure: Check of Calculation Method to Estimate Mass Flow Rate and Critical Flow Conditions" - Proceedings of the Seventh International Conference "Multiphase Flow in Industrial Plants" - Bologna, Italy, settembre 2000
M.Tardiola, - "Prove di Qualificazione di Valvole di Sicurezza in Biofase a Contropressioen Variabile" - Tesi di Laurea in Ingegneria Meccanica, Università degli Studi "La Sapienza" di Roma, ottobre 2001
R. Misiti, -"Studio dell'efflusso bifasico attraverso una valvola di sicurezza" - Tesi di Laurea in Ingegneria Chimica, Università degli Studi "La Sapienza" di Roma, marzo 2002
Boccardi G., Bubbico R., Celata G.P., M. Cumo- "Efflusso Bifase Attraverso Valvole di Sicurezza. Confronto tra i Dati Sperimentali e le Previsioni di Alcuni Modelli Omogenei" - Proceedings of thr 20th U.I.T. National Conference On Heat Transfer, MARATEA, Italy, June 27-30, 2002
Boccardi G., Bubbico R., Celata G.P., M. Cumo- "Water Two-Phase Flow through Pressure Safety Valve with Variable Backpressure: Check of Calculation Methods to Estimate Mass Flow Rate and Critical Flow Conditions" - Proceedings of the Eighth International Conference "Multiphase Flow in Industrial Plants" - Alba, Italy, settembre 2002

Top