Effect of salt content in the corrosion

8/10/2019 Effect of salt content in the corrosion

1/9

Nahrain University, College of Engineering Journal (NUCEJ) Vol. 13 No .1 , 2010 pp -66 -73

NUCEJ Vol .13, No .1 Salt Content on The Corrosion Rate 66

Effect of Salt Content on The Corrosion Rate of Steel Pipein Turbulently Flowing Solutions

Dr. Basim O. Hasan *

Chem. Eng. Dept Nahrain University -Baghdad -Iraq

Abstract The effect of salt content on the corrosion

rate of carbon steel pipe in aerated waterunder isothermal turbulent flow conditions(Reynolds number = 15000 -110000 andtemperature of 30-60 oC) is investigated byweight loss method. The test solutions weredistilled water (one time distilled) and sea

water (0.5N NaCl ). Also the results of 0.1 NNaCl solution obtained by Hasan [ 2003 ]were considered. The effect of Reynoldsnumber temperature on the corrosion rateis also studied and discussed. For each run,nine specimens of carbon steel were placedin the test section and the corrosion rate foreach one was determined at every Re andtemperature and the average was taken.The variation of friction factor with Re andtemperature was determined by measuringpressure drop across the test section.

The results revealed that the salt content

has considerable effect on the corrosion ratefor the whole investigated range of Re andtemperature. The highest corrosion rate wasencountered in sea water ( 3.5% NaCl ). Alsodistilled water causes considerable corrosionrate but it is lower than for the whole rangeof Re and temperature. Also it is found thatReynolds number increases the corrosionrate depending on temperature.Temperature affects the corrosion rate bychanging two main parameters, oxygensolubility and diffusivity.Key words: corrosion, salt, turbulent,

carbon steel, friction factor. Introduction

Most aqueous solutions (ranging from bulknatural water and chemical solutions to thincondensed films of moisture) will be in contactwith the atmosphere and will contain dissolvedoxygen which can act as a cathode reactant.The solubility of oxygen in water decreasessignificantly with the increase in temperatureand slightly with concentration of dissolvedsalts [Shrier 2004 ]. The rate of corrosionprocesses with oxygen depolarization is

determined mostly by the rate of oxygendiffusion to the metal surface. This kind ofcorrosion is extremely important, since it

includes practically important process such ascorrosion of iron and steel in neutral saltssolutions, corrosion of zinc in several neutralsolutions, various cases of copper corrosion[Akimov 1958, Moreira et . al. 2004 ].

In many corrosion problems, there isstrong evidence that the rate of uniformcorrosion is controlled by the rate of masstransfer. This is true whether the corrosionfluid remains static or in fast motion withrespect to the metal surface. However,molecu lar diffusion is not the only factorwhich influences the rate of corrosion. Inaddition, in turbulent fluids, the rate oftransport of eddy diffusion appears toparticipate in the control of the over- alltransfer rate. It is well known that the rate-contr olling step in most natural water corrosionprocesses is the cathodic half reaction ofoxygen reduction[Mahato et al . 1980, Moreiraet. al. 2004, Cheng and Steward 2004 ].

For reaction rate that is diffusioncontrolled, the limiting current density is :

iL= z F k C 1

Thus the basis of the LCD technique indetermining k is to measure LCD at fixed C.Also if k is known then measuring i L allowsthe determination of C, which will be thebulk concentration. If this approach can beapplied to a corrosion process in which both ananodic and cathodic reaction occur, if eitherreaction is diffusion controlled then the rate of

corrosion will be given by

ic= zFk C 2

The value of C will be the bulk concentrationof cathodic reactant for systems which arecathodically controlled and will equal thesolubility limit of the relevant product whenthe anodic reaction is diffusion controlled. It isfound from dimensional analysis that:

Sh= cons. Re x Sc y 3

x is usually between 0.3 and 1 , y is typically0.33 [Poulson 1983]. The geometry of a


2/9

NUCEJ Vol.13 No .1 Hasan 67

system, i.e., flat plates, stirrers, pipes, etc. havea large effect in determining the influence ofReynolds number on corrosion rate, i.e., x .

Chlorides have probability received mostconsideration in relation to their effect oncorrosion. The effect of sodium chloride

concentration on the corrosion of iron in airsaturated water at room temperature was foundto increase the corrosion rate. Chloridesincrease the electrical conductivity of the waterso that the flow of corrosion currents will befacilitated [Uhlig 1985, Shrier 2000]. Later,Kolman et . al. [2002 ] found the corrosionpotential becomes more negative withincreasing NaCl concentration . Bahar [2002 ]showed that, in stationary aerated water, thelimiting current density increases withincreasing NaCl concentration up to 3.5% , andthe corrosion potential becomes more negative.

Electrical conductivity is often a convenientand accurate measurement of salinity orchlorinity. Dissolved oxygen is a veryimportant factor in the corrosion of metalsimmersed in sea water . Table 1and 2 show theoxygen contents of different solutions.The aim of this work is to study the influenceof salt content of water for wide range ofReynolds number and temperature on thecorrosion rate carbon steel pipe under fullydeveloped turbulent flow conditions usingthree types of solutions, distilled water and seawater (0.5 N NaCl) , and the data of Hasan

[2003 ] for 0.1 N NaCl solution .

Table 1: Oxygen Solubility at AtmosphericPressure [Perry and Chilton 1998, Sense 2001].

T, oC Solubility, mg/l Distilled Water

Solubility,mg/l

0.1 N NaCl 30 35

40

45

50

60

7.559

6.625

6.401

6.001

5.477

4.501

7.500

6.95

6.389

5.990

5.399

4.490

Table 2: Oxygen Content of Sea Water [Shrier2000 ]

T, oC Solubility, mg/l (Sea water)

-205

10 15 20 30

8.5 8.08 7.16 6.44 5.86 5.38 5.42

Experimental Work

The dynamic system for performing theexperimental work is illustrated in Fig . 1. It iscomposed of: electrolyte reservoir, heater andcontroller, pump, rotameter, specimens, PH-meter, inverted U-tube manometer, controlvalves, PVC flanges, and PVC tubes. The testsection was composed of nine cylindricalspecimens each has an inner diameter of2.11 cm and outer diameter of 2.45 cm and 3 cmlong. The specimens were mounted in amanner to avoid discontinuities in between thespecimens. The test section is mounte dhorizontally and preceded by 170 cm entranceregion (i.e. , L/d=80 ) of PVC tube to ensurefully developed velocity profile before thesolution reaches the specimens [Bennett andMyers 1982 , Brodkey and Harshy 1989 ]. Thetest section is followed by 55 cm exit length toavoid disturbance at the outlet. Two staticpressure taps were installed at the edges of thetest section flash to the inside surface of theextension PVC pipe. The holes of pressure tapswere of 1.5 mm in diameter and the tapsinstalled flash to the inside surface to measurethe static pressure drop and avoid thedisturbance of velocity profile [Coulson andRichardson 1998 ]. Prior to use the specimenswere abraded in a sequence of emery papersgrades : 120 , 200, 320 , 400 , and 600 , washedby tap water followed by distilled water, driedwith clean tissue, immersed in analar methanolfor 5 min , rinsed with clean acetone , and driedwith clean tissue. The specimens then werestored in vacuum desicator over high activitysilica gel for 1.5 h before use, then they wereweighed using sensitive balance and directlyexposed to the corrosion environment .Thesolution with a particular Re and temperaturewas pumped to the test section for 6 hours andthe pressure drop across the later was takenafter 5 minutes from the test run start.

After exposure to the corrosion environmentthe specimen was washed by tap water thenbrushed by smooth brush under running tap


3/9


water to remove part of corrosion product thatis not strongly adherent. Then the outer surfac eof specimen is insulated by teflon tape and thespecimen was immersed in 20% warmhydrochloric acid containing organic inhibitor(hexamine ) for 1.5 minutes [Fontana and

Green 1994 ]. After that the specimen waswashed by tap water, distilled water, dried withclean tissue , kept in dissector for 1.5 h, andthen weighed. Blank tests showed noappreciable weight loss caused by the cleaningsolution.

Results And DiscussionFriction Factor

The Friction factor across the test sectionwas determined by measuring the pressuredrop via inverted U -tube manometer. where

LuPd

f 224

Over the years, friction factor has receiveda wide attention from authors. Numerousstudies have been performed to study factorsinfluencing friction factor under differentconditions, e.g., isothermal and heat transferconditions [Colburn 1933 , Mody 1944 , Smithand Epstein 1957 , Knudson and Katz 1958 ,Petukhov 1970 ] and roughness conditions

[Mody 1944 , Knudson and Katz 1958 ]. Mostof these studies have showed that the maintwo variables influencing friction factor ( f )were Re and surface roughness. The variationof friction factor with Re for clean surface (nocorrosion products formed) at differenttemperatures is presented in Fig.2 . It isobvious that f decreases with Re for alltemperatures. Also the figure reveals noappreciable effect of temperature on thefriction factor. No noticeable differencebetween friction factor of distilled water andsea water is noticed.

Corrosion Rate

Effect of salt on the corrosion rate

Figures 3 to 5 show the variation ofcorrosion rate expressed in gmd (gm/m 2.day)with Re at different temperature for the threesolutions. It is evident that increasing Re,increases the corrosion rate for the whole rangeof temperature. The sea water causes highestcorrosion rate and distilled water causes thelowest corrosion rate, i.e, the higher the saltcontent is the higher the corrosion rate. This isascribed to the increased solution electricalconductivity of the solution facilitating the

cathodic reaction (oxygen reduction) andanodic reaction (iron dissolution). I is to benoticed that the distilled water causesconsiderable corrosion rate for the wholeinvestigated range of Re and temperature. Inpresent work the occurrence on corrosion in

distilled water may be attributed to thepresence of little amount of salts in the distilledwater because it is one time distilled. It is to benoted from Figs . 3 to 5 that the increase incorrosion rate with NaCl concentration in Fig.3 (for T=30 oC) is lower than in case of 45 oCand 60 0C. This is ascribed to the fact thatincreasing temperature increases the electricalconductivity of the solution as evidenced byprevious work [Shrier 2000].

Effect of Reynolds Number

Fig. 3 to 5 shows that at a particulartemperature the corrosion rate increases withRe for the three solutions. Since the corrosionof iron in aerated water is a mass transfercontrolled process [Foroulis, 1979; Mahato etal, 1980 ; Smith et al , 1989 ; Shreir , 2000 ],therefore increasing Re (or velocity) willincrease the amount of oxygen arriving to thesurface and hence the corrosion rate isincreased. Increasing Re leads to decrease thethickness of diffusion layer in the wall vicinitythat represents the main resistance to oxygentransport [Brodkey and Hershey 1989 , and

Thomas 2000 , Welty et al 2001


4/9


No. Name No. Name

1 0.1 N NaCl Solution 8 PVC Flanges

2 Thermometer 9 Inverted Manometer

3 PH Meter 10 Venting Valve

4 Heater & Controller 11 Rotameter

5 Connection Pipes 12

Test section Valve

6 Specimens 13 Centerfugal Pump

7 Fixation Screw 14 By Pass Valve

Fig .1: Experimental Apparatus

2

346 7 8910

12

11

14

5

13


5/9


Effect of TemperatureObserving Fig. 6 , for distilled water ,

shows that at high Re the higher thetemperature is the higher the corrosion rate.While at low Re, the corrosion rates in threetemperatures are close. The same trend holdsfor sea water. This behavior can be explainedas follows: increasing temperature leads tochange two variables that act in a conflictingway. Firstly, increasing temperatureaccelerates the reaction rate and diffusion rateof oxygen species by increasing the moleculardiffusion coefficient. Secondly, as thetem perature increases the oxygen solubilitydecreases [Sheir 1994]. Third factor will playan important role in determining the corrosionrate is Re (or velocity). The increase in Reenhances the eddy diffusion of O 2 from thebulk of solution to the metal surface whichincreases O 2 transport to the metal surfaceleading to higher corrosion rate. Strictlyspeaking, the combined effect of threevariables, namely, oxygen solubility, oxygenmolecular diffusivity, and solution velocity (oreddy diffusion), will determine the trend ofcorrosion rate with increasing temperature. The results revealed that the position of thespecimen in the test section has no influenceon its corrosion rate, i.e, no entrance effect ofmass transfer was noticed, that is in harmonywit h previous observations [Mahato et al1980 ]. This may be ascribed to the formedcorrosion product layer which may disturb theuniformity of active sites and the oxygenamount near the surface by increasing theturbulence. Also the difference in corrosionrate among the nine specimens reaches 5% .

Fig. 2: Variation of friction factor with Re atdifferent temperatures.

Fig. 3: Variation of Corrosion Rate with Re forVarious Solutions at 30 oC.

Fig .4: Variation of Corrosion Rate with Re forVarious Solutions at 45 oC.

8 9 2 3 4 5 6 7 8 9

10000 100000

Re

6

7

8

9

10.00

f x

1 0

Distilled Water

T=30 C

T=45 C

T=60 C

- 3

15000 30000 45000 60000 75000Re

40

80

120

160

200

C o r r o s i o n R a t e ( g m d )

T=30 C

Sea Water

0.1N Nacl Solusion

Distilled Water

0 20000 40000 60000 80000 100000

50

100

150

200

250

300

350

C o r r o s i o n

R a

t e ( g m

d )

T=45 C

Sea Water

0.1N NaCl Solution [Hasan 2003]

Distilled Water


6/9


Fig. 5: Variation of Corrosion Rate with Re forVarious Solutions at 60 oC.

Fig . 6: Effect of Temperature on the CorrosionRate.

Conclusions

From present investigated system thefollowing points can be concluded:1. Increasing salt content increases leads toincrease the corrosion rate for the entireinvestigated range of Re and temperature byincreasing the solution electrical conductivity. 2- Increasing Re, increases the corrosion ratefor the three solutions at all temperatures. 3. Increasing the temperature affect thecorrosion rate by changing two parametersacting in a conflicting way, i.e, oxygensolubility and oxygen diffusivity. Alsoincreasing temperature increases the solutionelectrical conductivity enhancing the corrosionrate.

Nomenclature

C Concentration, mole/m 3.

d Pipe diameter, m.

FFaradays constant ( 96487 Columb/equivelent)

f Friction factor.

i Current density, A/m2

.

k Mass Transfer Coefficient, m/s.

L Length, m

Re Reynolds number= du/

Sc Scmhidt number= / D

Sh Sherwood number= kd/D.

u Velocity, m/s

p Pressure drop, N/m 2 .

Viscosity, kg/m.s 2.

Kinematic viscosity , m 2 /s.

20000 40000 60000 80000 100000 120000

Re

0

100

200

300

400

500

C o r r o s

i o n

R a t e

( g m

d )

T=60 C

Sea Water

0.1N NaCl Solution [Hasan 2003]

Distilled Water

0 20000 40000 60000 80000 100000 120000

Re

100

200

300

C o r r o s

i o n

R a t e

( g m

d )

Distilled Water

T=30 C

T=45 C

T=60 C


7/9


Abbreviations

gmd Gram/m 2. day

L Limiting

c Corrosion

References

1-Akimov, G. V., Corrosion- NACE , Vol .14,P. 463 , 1958 .

2-Bahar S. S., M. Sc. Thesis, Chem. Eng.Dept ., Saddam University , Bahgdad , 2002 .3-Bennett C.O. and J. E. Myers, Momentum,Heat and Mass Transfer , 3rd Edition, Mc GrawHill , United States , 1982 .

4-Coulson J. M. and J. F. Richardson,Chemical engineering , Vol . 1,5 th Edition,Butter Worth Heinemann , Britain , 1998 .

5-Fontana M. G., N. D. Greene, CorrosionEngineering , 2nd st Edition.,London , 1984 .

6-Foroulis, Z. A., "The Influence of Velocityand Dissolved Oxygen on the Initial CorrosionBehavior of Iron in High Purity Water",Corrosion , Vol. 35 , P. 340 -344 , 1979 .

7-Knudsen J. G. and D. L. Katz, FluidDynamics and Heat Transfer, Mc Graw Hill,New York , 1958 .

8-Kolman D. G., D. K. Ford, D. P. Butt, and T.O. Nelson, Materials Corrosion andEnvironmental Effects Laboratory, LosAlamos , NM 87545 , 2002 .

9- Mahato B. K., C. Y. Cha, and W. Shemlit,Corros. Sci ., Vol . 20, P. 421 , 1980 .

10-Nikur dse J., VDI- Forschungsheft, 356,1932

11 -Perry R.H. and C.H. Chilton, Chem Eng.Hand -Book, 6th Ed., Mc Graw Hill , 1998 .

12-Perry R.H. and C.H. Chilton, Chem Eng.Hand -Book, 6th Ed., Mc Graw Hill , 1998 .

13-Petukhov B.S. , Heat Transfer and Frictionin Turbulent Pipe Flow with Variable PhysicalProperties , Advances in Heat Trans. , Vol . 6, P.503 , 1970 .

14- Sense F., Oxygen Solubility, NorthCalifornia State , 2001 .

15-Shreir L . L., Corrosion Handbook , 3nd

Edition ., Part 1, Newnes -Butter., London,2000.

16-Smith S. W., K. M. McCabe, and D.W.Black, " Effect of Flow Parameters on theCathodic Protection of a Steel Plate in NaturalSeawater", Corrosion- NACE, Vol. 45, No . 1,P. 790 (1989 ).

17-Ulig, H.H., Corrosion and CorrosionControl , 3rd Edition., Wiley-I ntersciencePublicatio , Newyork, 1985 .

18- W. J. Thomson, Introduction to TransportPhenomena , 1st Edition, Prentice Hall PTR,New Jersey, 2000

19-. Welty, C. E. Wicks, and G. Rorrer,

Fundamentals of Momentum, Heat, and MassTransfer , 4th Edition, John Wiley and Sons,United States of America , 2001

20- Brodkey and H. C. Hershey, TransportPhenomena , 2nd Printing, Mc Graw Hill, NewYork, 1989 .

21- Poulson B ., Corrosion Science , Vol , 23,No . 1,


8/9


0.5 N

NaCl0.1 NNaCl

0.5 N


9/9

This document was created with Win2PDF available at http://www.daneprairie.com.The unregistered version of Win2PDF is for evaluation or non-commercial use only.
http://www.daneprairie.com/http://www.daneprairie.com/http://www.daneprairie.com/http://www.daneprairie.com/

Effect of salt content in the corrosion

Documents

Transcript of Effect of salt content in the corrosion