# CALCULATION OF MASS TRANSFER IN MULTIPHASE OF MASS TRANSFER IN MULTHWASE FLOW L. WANG, M. GOPAL NSF,...

date post

14-May-2018Category

## Documents

view

214download

0

Embed Size (px)

### Transcript of CALCULATION OF MASS TRANSFER IN MULTIPHASE OF MASS TRANSFER IN MULTHWASE FLOW L. WANG, M. GOPAL NSF,...

Paper Nom

50

CALCULATION OF MASS TRANSFER IN MULTHWASE FLOW

L. WANG, M. GOPAL

NSF, I/CJCRCCORROSION IN MULTIPHASE SYSTEMS CENTER

DEPARTMENT OF CHEMICAL ENGINEERING

OHIO UNIVERSITY, ATHENS, OHIO, USA

ABSTRACT

This paper summarizes the results of mass transfer mechanisms under disturbed liquid-gas flow in10 cm diameter pipe using electrochemical limiting current density and potentiostatic noise technique.The solution used is potassium ferro/ferricyanide dissolve in 1.3 N sodium hydroxide system. Mass

transfer coefficients in full pipe flow and slug flow are obtained. The relationship between mass transfercoefficient with full pipe flow velocities and with slug flow Froude numbers are studied. The impact ofbubbles in slugs on the mass transfer coefficient is revealed, The impact of flow disturbance, includingweld beads and pits, are discussed for both fhll pipe flow and slug flow.

Keywords: full pipe flow, slug flow, disturbed flow, mass transfer, limiting current method,polarization, corrosion.

INTRODUCTION

Slug flow exists in the pipelines at high production rate of oil and gas. High velocity slugs are veryturbulent and the corrosion rate is increased greatly in this flow regime. This is due to pulses of gasbubbles entrained in the mixing zone being forced towards the bottom of the pipe. There they impact andcan collapse, causing localized corrosion.

Copyright@l 998 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole must be made in writing to NACEInternational, Conferences Division, P.O. Box 218340, Houston, Texas 77218-8340. The material presented and the views expressed in thispaper are solely those of the author(s) and are not necessarily endorsed by the Association. Printed in the U.S.A.

Further, severe corrosion is found in the vicinity of flow disturbances such as weld beads, pits,and pipe connections. Thk is thought to be due to the enhanced turbulence and mass transfer rate aroundsuch obstacles. No information exists at the present time regarding the mass transfer mechanisms underdisturbed flow conditions.

It has been found that the corrosion rate is strongly related to the mass transfer rate of corrosiveions in the multiphase mixture from the bulk to the pipe wall. It is suggested that the mass transfercoefficient is the main parameter in the corrosion modeling (Zhang et al., 1997). However, the masstransfer coetlcients in large diameter pipe flow and in multiphase flow are not known. It would be veryhelpful to measure mass transfer coefficients for multiphase flow systems of interest in oil and gasproduction.

Mass transfer measurements using limiting current method provide a convenient and accuratemeans to determine the local mass transfer coefficient. Small insulated electrodes embedded in flow pipewall could provide local mass transfer coefficients. Due to their fast response, instantaneous fluctuatingvalues, which directly reveal the local turbulence could be obtained by potentiostatic current noisemeasurement.

Reiss (1962) described in detail, the technique of making measurements of mass transfer using a

diflhsion controlled electrolytic reaction, and the use of this tectilque to measure velocity fluctuations atthe wall. Measurements of mass transfer coefficient and intensity were carried out in 2.54 cm diameterpipe using the potassium ferro-ferricyanide electrochemical system. The electrodes were made from threesizes of nickel wire (1.636 mm, 0.664 mm, 0.397 mm in diameter respectively).

Shaw andHanratty(1964) measured local time average mass transfer coefficients with embeddedtest electrode. Their experiment differed from that of Reiss in that the test electrodes, instead of beingsurrounded by inert surface, are surrounded by active electrode surface. Test electrodes of four differentdiameter were used, They found the very large magnitude of the fluctuations in the mass transfer rate(mass transfer intensity is as much as 0.47), the relative sizes of the longitudinal and circumferentialscales, and the low frequency scales of mass transfer fluctuations.

Sirkar and Hanratty (1970) obtained the root-mean-square fluctuating mass transfer coefficientfor a Schmidt number of about 2300 in a 7.62 cm diameter pipe, They used order-of-magnitude analysisand concluded that flow fluctuations in the direction of mean flow have little effect on the mass transferfluctuations. The local value of mass transfer coefficient they obtained is more accurate compared withShaw & Hanratty.

Mizushina (1971) discussed the method of diffhsion-controlled electrochemical reaction and itsapplications in the study of transport phenomena, including mass transfer measurements, shear stressmeasurements and fluid velocity measurements. He also provides a review emphasizing the application oflimiting current measurements on microelectrodes for local and instantaneous shear stress and velocitydeterminations, as described by Hanratty (1966).

5012

Selman and Tobias (1978) compiled over a hundred mass transport correlations pertaining todifferent transport, flow, and agitation configurations. The theory and practice of limiting-currenttechnique for the measurement of mass-transport coefficients were described.

Landau (198 1) used limiting current technique to determine the mass transport rates of laminarand turbulent flows. He also discussed the consideration of current distribution, selection and placementof the electrodes, modes of current application and the electrochemical system selected.

Campbell and Hanratty (1983) studied the structure of the velocity field close to a wall bymeasuring the transverse component of the fluctuating velocity gradient simultaneously at multiplelocations on the wail. Velocity fluctuations of all frequencies appear to have the same transverse scale.Measurement were held in 8 inches pipe system. It is found that frequency spectra and intensities areuniversal properties and not strongly rdTected by the particular design of the experimental flow system.

Mass transfer coefficient is calculated from the average limiting current using the followingequation:

K= I~/(n FACJ (1)

where K = mass transfer coefficient1~= limiting currentn = number of moles reactedF= Faradays constantA = surface area of the electrodecb = bulk concentration of the potassium ferrocyanide

Mass transfer deviation is calculated from the potential static noise curve using the followingequation:

DevK = [.Z(k - Avg~2/N]l2 (2)

where Dev~ = mass transfer deviationk = instantaneous mass transfer coefficientAvg~ = mean value of the instantaneous mass transfer coefficientsN= total number of the instantaneous mass transfer coefficients

EXPERIMENTAL SETUP

The overall layout of the system is shown in Figure 1. The flow loop is a 10 m long 10 cm indiameter Plexiglass pipe. A 1 m3 stainless steel tank is filled with 1,3 N sodium hydroxide and 0.01 Mpotassium ferricyanidelferrocyanide. Nitrogen is stored in a pressured gas tank and is added into thesystem through a pressurized regulator and a needle valve.

5013

The test section is a 10 cm diameter Plexiglas pipe with electrodes and disturbance block. Theorientation of the electrodes in the direction of flow are as follows: the reference electrode, disturbanceblock, working electrode and counter electrode. The overall layout of it is shown in Figure 2. Thedisturbance block simulates the disturbance inside a pipe line, such as weld bead and pits. The weld beadis a small hump. The pit is a small cylinder hole. The counter electrode is a ring electrode mounted flushwith the pipe wall. The working and reference electrodes are hastalloy pins inserted in the Plexiglassblock, laying out equal-distantly in a line at the bottom of the pipe. The distance between two consecutiveelectrodes is 4.5 mm and the surface area of it is 7.85x 10-7 square meter. The dimensions of the weldbeads and the pits are described in Table 1 and 2.

The data were taken by Gamry soflware CMS 100 installed in a Pentium computer. Two typesof results were collected. The first was the DC polarization curve, from which the anodic limitingcurrent was determined. The other was electrochemical current noise curve that was measured by settingthe potential value at which the limiting current was reached.

Table 1, The Dimension of the Weld Bead Test Section

Test Section 1 mm Weld Bead 2 mm Weld BeadHeight (mm) 1 2

Table 2. The Dimension of the Pit Test Section

Test Section Pit #1 Pit #2 Pit #3 Pit #4 Pit #5 Pit #6Diameter (mm) 2 4 2 4 2 4Depth (mm) 2 2 4 4 6 6

EXPERIMENTAL TEST MATRIX

Full pipe flow and slug flow experiments were conducted for the ferrolferncyanide system. Inorder to study the effect of disturbance resource, nine test sections were studied at three differentvelocities. Table 3 showed the Experiments carried out.

Table 3. Experimental Test Matrix

Flow Velocity (for fill pipe flow, m/s) 0.5, 1, 2

Froude Number (for slug flow) 4, 6, 9

Test Section 2 Weld Bead, 6 Pit, 1 Smooth Section

Electrode 1,2, 3,4, 5

5014

RESULTS AND DISCUSSION

Full Pipe F1OWResults

The results of smooth test section, pit #1 and 2 mm weld bead section are shown in Figure 3.When the flow velocity is 0.5 mls, the average mass transfer coefficients in these three test sections are0.83 x10-4 m/s, 0,96x 10-4 rds and 1,00x 10-4rds respectively, As the flow velocity increases to 2.0 m/s,the mass transfer coefficients increase to 1.72 x10-4m/s, 2.43 x10-4 rds and 3.12 x10-4mls respectively. It isseen that the mass transfer coefficients increase with increase of full pipe flow velocity in disturbed andundisturbed flow. At the same i-low velocity, such as Imls, the ma

Recommended

*View more*