Corrosión en Plantas de GN

7/23/2019 Corrosión en Plantas de GN

http://slidepdf.com/reader/full/corrosion-en-plantas-de-gn 1/7

The interaction of mercury and aluminium in heat

exchangers in a natural gas plants

R. Coade *, D. Coldham

HRL Technology Pty Ltd, 677 Springvale Road, Mulgrave, Vic. 3170, Australia

Abstract

This paper reviews current understanding of mercury induced liquid metal embrittlement (LME) and the mechanism of failure in aluminium

heat exchangers. Natural gas can be contaminated with low levels of mercury, which can concentrate in cryogenic heat exchangers. There havebeen several instances where LME has led to major failures and gas leakage in gas processing plant.

q 2006 Elsevier Ltd. All rights reserved.

Keywords: Liquid metal embrittlement; Mercury; Aluminium heat exchangers; Natural gas

1. Introduction

Liquid metal embrittlement (LME) is a complex metal

fracture mechanism that occurs without warning. Liquid

mercury has been known to have a potentially devastating

effect on aluminium for several decades [1–4], and the

accumulation of mercury in parts of natural gas plants hasled to failures. In the 1970s, LME and corrosion of aluminium

alloys by mercury in liquid natural gas (LNG) industry became

a cause for concern. A paper published in 1980 entitled

‘Mercury—LNG’S Problem’ [5] focussed on the potential risk

of failures and papers published in Proceedings of GPA Annual

Conventions and elsewhere in 1990s discussed failures of

components due to LME by mercury and on methods to combat

this [6–8].

The conditions required for LME to occur are:

(1) the presence of an embrittling liquid metal—with mercury

being a well-known, severe embrittling agent foraluminium alloys,

(2) the presence of a stress above a threshold value, which can

be as low as 5% of the yield stress under some conditions

for aluminium alloys in mercury, and

(3) ‘wetting’ of the substrate by the liquid metal, which in the

case of aluminium alloys requires rupture of the oxide film

between the substrate and liquid metal.

2. Natural gas processing to produce liquids

The natural gas purchased by consumers consists almost

entirely of methane, the simplest hydrocarbon. In gas

reservoirs, however, methane is typically found with heavier

hydrocarbons—such as ethane, propane, butane and pentane.

The raw gas also contains water vapour, hydrogen sulphide,carbon dioxide, nitrogen and other gases that are removed from

the gas stream at processing plants.

In gas processing plants, hydrocarbons are separated

through fractionation—based on the different boiling points

of the hydrocarbons in the natural gas liquids (NGL) stream.

The liquids are cooled to temperatures aroundK50 8C and the

various fractions are separated as they boil off as the liquids

temperature is increased in stages in various heat exchangers.

This cryogenic distillation, separating ethane and heavier

hydrocarbons from sales gas (methane) occurs within cold

boxes, typically made from aluminium. An example of such a

coldbox is shown in Fig. 1.

3. Forms of Hg attack in aluminium heat exchangers

Mercury can occur in natural gas feed stock, often at very

low levels, and can sometimes accumulate in quantities

sufficient to cause severe attack and failure of cryogenic

aluminium heat exchangers. The cooling equipment in a gas

separation process is typically an aluminium plate-fin heat

exchanger, the construction of which is often an Al 3003 core

with Al 5083 or 6061 headers, nozzles and piping. The mercury

in the natural gas can degrade the aluminium coldbox materials

by three basic mechanisms [6]:

International Journal of Pressure Vessels and Piping 83 (2006) 336–342

www.elsevier.com/locate/ijpvp

0308-0161/$ - see front matter q 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.ijpvp.2006.02.022

* Corresponding author.

E-mail address: [email protected] (R. Coade).

http://www.elsevier.com/locate/ijpvp

mailto:[email protected]

mailto:[email protected]

http://www.elsevier.com/locate/ijpvp



3.1. Amalgamation

Amalgamation is the process by which mercury forms liquid

solutions with various metals, primarily Al, Au, Ag and Zn. In

the case of aluminium, the concentration of aluminium in the

amalgam is low and thus the depth of attack is limited.

Furthermore, aluminium is generally prevented from contact

with mercury by the Al2O3 protective surface oxide—for the

amalgam reaction the mercury must wet the aluminium metal

surface. The oxide on aluminium is not homogeneous and

contains numerous defects, but in general mercury will notmigrate through these microscopic cracks and defects to reach

the underlying metal. However, if the extent or severity of such

defects is increased by thermal or mechanical stresses, abrasion

or some chemical environments there is increased risk of

mercury damage.

3.2. Amalgam corrosion

Amalgam corrosion is the combined action of mercury and

moisture producing a corrosion process that propagates with

miniscule amounts of mercury. The reaction is:

HgCAl/HgðAlÞ amalgam (1)

HgðAlÞC6H2O/Al2O3$3H2OC3H2CHg (2)

Small amounts of aluminium can dissolve in liquid mercury,

diffuse to the mercury–moist air interface, and then rapidly

oxidize. Since, oxidation removes aluminium from the

mercury, further aluminium can dissolve, and the process can

continue until the aluminium is completely converted to oxide.

In practice, voluminous oxide whiskers and deep pits are

observed. Rapid oxidation requires the presence of moisture,

and reaction rates are slow in its absence.

Gordon [9] citing the early work of Pinnel and Bennett

[10,11] and Bruce and Wise [12] described the amalgamate of

Al and Hg as a white fibrous corrosion product identified as

either an amorphous form of g-Al2O3 or a form of hydrated

Al2O3 (alumina) or alumina hydroxide. Phannensteil et al. [13]

concluded that it was necessary for ions to be present in the

water for amalgam corrosion to occur. It is probable that theseions destabilised the protective oxide layer allowing the

mercury to come in contact with the base metal.

Nelson [8] acknowledges that mercury can cause cata-

strophic attack of aluminium in the presence of ‘free’ water,

however, suggests that in cryogenic heat exchangers the

presence of water is extremely unusual, and has not been a

factor in any reported leak occurrence.

3.3. Liquid metal embrittlement (LME)

LME of Al alloys by mercury is one example of a generic

phenomenon in which many (but not all) metals are embrittledby certain liquid metals [14]. For example, mercury embrittles

not only Al alloys but also Cu, Ti, Ni, Fe, and Zn alloys, but

does not embrittle Mg alloys [15]. Al alloys are also embrittled

by, for example, liquid Ga, In, Pb, Sn, Cd, and Na. There are

several types of LME (with different mechanisms), but most

cases involve only adsorption of embrittling atoms at stressed

surfaces and crack tips, i.e. no diffusion of embrittling atoms

into the material or ahead of crack tips is involved [16,17]. The

Al:Hg system falls into this category and, hence, other types of

LME are not considered here.

LME is generally much more severe than other embrittling

processes, such as hydrogen-embrittlement or stress-corrosioncracking, and once cracks have initiated, very rapid sub-critical

cracking can occur even at low stresses (stress-intensity

factors) [18]. Cracking occurs preferentially along grain

boundaries for the Al:Hg couple (and for many other couples),

but transgranular (cleavage-like) fractures can also occur.

Liquid metals are drawn into growing cracks so that the crack

tip is always in contact with embrittling metal atoms. (The rate-

controlling process for cracking is still being debated, but the

rate of flow of liquid within cracks may control the rate of

cracking in some circumstances.)

Adsorption of embrittling atoms at crack tips weakens

substrate interatomic bonds, and facilitates crack growth by

enabling interatomic bonds to break or shear more easily than

in inert environments. Preferential adsorption at grain-

boundary/surface intersections results in the preference for

intergranular fracture. Much less plasticity is associated with

fracture in liquid metal environments than in air, and fracture

surfaces can be featureless or can exhibit small, shallow

dimples, whereas large deep dimples are produced by fracture

in air. Thin films of liquid metal are left behind the advancing

crack tip and, hence, fracture surfaces are covered with a film

of liquid metal. For the Al:Hg system, ‘de-wetting’ can occur

so that small globules of mercury are present on fracture

surfaces. The presence of mercury on fracture surfaces can also

result in the growth of oxide whiskers after fracture—a

(7)

(6)

Fig. 1. Schematic view of cryogenic heat exchanger showing the manifolds (6)

and nozzles (7).

R. Coade, D. Coldham / International Journal of Pressure Vessels and Piping 83 (2006) 336–342 337



phenomenon peculiar to aluminium and discussed in the

preceding section.

For LME cracks to initiate there must be intimate contact

between liquid and solid metals, with no intervening oxide

films to prevent wetting and adsorption. Al alloys are covered

by a thin, protective oxide film, and surfaces can be covered by

liquid mercury indefinitely without any reaction until the oxideis damaged. Oxide films can be broken by mechanical

processes, e.g. by scribing or abrasion, by chemical processes,

e.g. corrosion, or by plastic deformation of the aluminium

resulting in slip steps at the surface. Slip processes can occur

locally in favourable oriented grains at stresses well below the

macroscopic yield stress, and slip steps can emerge at the

surface after long times under stress due to creep processes

(thermally activated dislocation re-arrangements) beneath the

surface.

The time required for slip-step/oxide-rupture events will, of

course, decrease with increasing stress and, hence, times to

LME initiation will also decrease with increasing stress. The

kinetics of LME crack initiation and growth depend not only on

stress (or stress-intensity factor) but also on many other

variables such as the composition of the liquid, the amount of

liquid, the composition, microstructure, and strength of the

substrate alloy and temperature [19].

The amount of liquid metal is important for several reasons.

Firstly, increasing surface coverage of the substrate by the

liquid metal could decrease the time to crack initiation, since

the probability of potentially weak sites in oxide films exposed

to the liquid would be increased. Secondly (and more

importantly), cracks can ‘run out of’ liquid metal if there is

only a limited supply because films of liquid metal are left on

fracture surfaces behind crack tips. Crack-arrest can thereforeoccur providing the stress-intensity factor, K , is below the

critical K for fast fracture in the absence of environmental

effects.

Cryogenic heat exchangers are often manufactured from

aluminium alloy 5083, an aluminium–magnesium alloy.

Magnesium silicide can contribute to age-hardening in these

alloys. At room temperature, aluminium can hold w1% Mg in

solution although heat exchanger alloys typically contain 4.5%

Mg. The aluminium rolling mills anneal this material at

w455 8C to dissolve all the Mg. Then the rolled stock is

quenched to room temperature to hold the magnesium in solid

solution. This results in a metastable solid solution that wants

to precipitate Al3Mg2. The kinetics of precipitation is so slow

at room and cryogenic temperatures, that for all practical

purposes, the alloy is stable.

However, when the alloy is welded, the temperature of the

heat affected zone can facilitate the precipitation of Al3Mg2.

This can result in a continuous or semicontinuous film of

Al3Mg2 being precipitated at the grain boundaries. The weld

structure is also subject to grain boundary precipitation because

of the solidification pattern and reheat from multi-pass welds.

Alloy composition and ‘temper’ can affect LME through

effects on strength, and grain-boundary microstructure (and

grain-boundary composition if segregation occurs), which can

affect creep rates and strain localization (hence, oxide-rupture),

and adsorption kinetics. Alloy strength is particularly

important, and increasing strength can either decrease or

increase susceptibility to LME depending on the circum-

stances. Crack growth rates generally increase, and threshold K

values for cracking generally decrease, with increasing

strength [20].

Specific data for welded 5083-0 alloy in liquid mercury at20 8C show that threshold K values are somewhat lower for

5183 weld material with an equiaxed microstructure than for

the 5083 plate (in the T-L orientation), although the crack

growth rate was somewhat higher for the latter for some K

values [21,22]. Limited data on the times to failure for the

welded 5083-0 alloy stressed in mercury at 20 8C show that

failure can occur in less than 1 h at high stresses (70% of yield),

and that there is considerable scatter with other similarly

stressed specimens lasting for w20 h. Failures occurred after

hundreds of hours at low stresses, with the threshold stress for

cracking being perhaps only w10% of the yield stress.

4. Breaching the aluminium protective oxide

For LME cracks to initiate there must be intimate contact

between liquid and solid metals, with no intervening oxide

films to prevent wetting and adsorption. Al alloys are covered

by a thin, protective oxide film, and surfaces can be covered by

liquid mercury indefinitely without any reaction until the oxide

is damaged. Oxide films can be broken by mechanical

processes, e.g. by scribing or abrasion, by chemical processes,

e.g. corrosion, or by deformation of the aluminium resulting in

slip steps at the surface.

In LNG plant, Gordon [9] favours the abrasion on the

surface by hard particles in the gas or liquid streams as the keyoxide breaching mechanism. He suggests that as the gas

entering cryogenic equipment usually consists of mainly

methane, CO2 and hydrogen, it is oxygen-free. Hence, the

reducing atmosphere of the gas stream may prevent any

reformation of the protective oxide layer once damaged from

hard particles in the gas stream has occurred.

Others suggest that the differential thermal expansion

between the aluminium substrate and the alumina oxide

being a factor of around 3 could cause the oxide to crack

when the heat exchangers is warmed.

It is known that slip processes can occur locally in

favourably oriented grains at stresses well below the

macroscopic yield stress, and slip steps can emerge at the

surface after long times under stress due to creep processes

(thermally activated dislocation re-arrangements) beneath the

surface. It is possible that such activity could lead to breaks in

the oxide film. The time required for slip-step/oxide-rupture

events will, of course, decrease with increasing stress and,

hence, times to LME initiation will also decrease with

increasing stress.

5. Metallography of LME cracking in heat exchangers

The coldbox on a cryogenic heat exchanger typically

includes a distribution manifold with several nozzles from

R. Coade, D. Coldham / International Journal of Pressure Vessels and Piping 83 (2006) 336–342338



the inlet manifold connected via headers to the heat exchanger

core. The manifolds are manufactured from seam welded pipe

and circumferential welds attach flanges. LME can lead to

extensive delamination cracking in the manifold itself, and also

cause extensive intergranular cracking in the field circumfer-

ential welds [23]. Some typical examples of such damage are

considered in the following paragraphs.In one case, LME led to cracking and failure in the manifold

just downstream from the flange, and involved cracking along

the longitudinal axis of the pipe at about the 6 o’clock position

for about 300 mm. There was an exceptional amount of

secondary cracking in many directions leading to the

detachment of substantial pieces of material and to major

delaminations parallel to the surface of the pipe. LME cracking

also ran around the circumferential weld area.

Examination of fracture surfaces showed that (i) there were

no shear lips or other macroscopic signs of ductility for any of

the cracks, (ii) all fracture surfaces were intergranular, and (iii)

all fracture surfaces were contaminated with small globules of liquid mercury and oxide whiskers characteristic of mercury-

wetted aluminium surfaces.

The longitudinal fracture surface was textured and directional,

with numerous ‘delamination’ cracks running normal to the

fracture surface, as shown in Fig. 2. This cracking extended for

some distance, in some areas greater than 50 mm deep, and the

crackingbecamebranched, as shown in the micrographs in Fig.3.

The length of the delamination cracking varied withposition. The cracks could deviate from the delamination

Fig. 2. Photograph showing the surface of the longitudinal crack and the

delamination cracks, which are more visible in the macroscopic view of a

section through the primary crack after polishing showing the nature and length

of cracking.

Fig. 3. Section showing multiple cracking near the surface of the longitudinal

fracture and near the end of the delamination crack. 165!.

Fig. 4. Examples of delamination cracking that has deviated and intersected the

internal surface, with alumina whiskers clearly exposing the cracks.




plane and intersect the internal surface of the vessel. In this

case, as the cracks were exposed to moist air they become

delineated with aluminium oxide whiskers, as can be seen in

Fig. 4

All cracks and fracture surfaces were found to be

contaminated with mercury and/or oxide whiskers consistent

with mercury induced corrosion. The presence of mercury on

the fracture surfaces, the intergranular nature of cracking and

the multiplicity of the cracking suggested a liquid metal

embrittlement phenomenon.

The circumferential weld metal is equiaxed and appears to

be more susceptible to intergranular cracking than does the

parent material in the manifold. This could be accentuated by

the presence of grain boundary precipitation although it is more

likely to be a result of a structure that presents fewer barriers to

crack movement than the heavily textured manifold. Secondary

cracking within the circumferential weld is typically branched,

following the equiaxed grain boundaries. An example of cracks

within the circumferential weld is shown macroscopically in

Fig. 5.The very branched nature of the cracking associated with

the circumferential weld is shown in Fig. 6, indicating the

equiaxed grain structure of the weld. The section through the

weld perpendicular to the longitudinal crack path shows

the growth of secondary cracks from the primary crack face

extending towards the outer surface, indicating a growth of the

primary fracture from the inner to the outer surface.

6. Fractography

Aluminium fracture surfaces can be cleaned in concentrated

nitric acid to remove corrosion product, contaminant and

mercury without damaging the surface. After cleaning the

surfaces could be examined in the scanning electron

microscope. Although traces of corrosion product remained,

key features of the fracture could be resolved.

The surface of the fracture through the circumferential weld

revealed a grain structure and some intergranular cracking as

can be seen in Fig. 7. It can be seen that although there is no

indication of extensive deformation, as would be expected in

intergranular LME cracking, there is micro-ductility as

indicated by the localised dimpled appearance across the

Fig. 7. Fracture surface of cracking in the circumferential weld showing the

intergranular and branched nature of cracks.

Fig. 5. Macroscopic view of a section through the circumferential weld

showing extensive branched intergranular cracking.

Fig. 6. Cracks at the outer surface of the circumferential weld and within the

weld showing extensive branched intergranular cracking. 100!95/141!.




grain face. The cracking had a similar intergranular appearance

at the outer surface in this region also but this was perhaps not

as pronounced as indicated in the photographs at the IDsurface.

The longitudinal crack in the manifold extended for nearly

300 mm from the circumferential weld, and various sections of

the longitudinal fracture surface were examined to identify any

differences in the fracture mode along the failure path.

Fig. 8 shows the appearance at the top of one of the ridges

mid-way through the thickness. The higher magnification

views show some micro-ductility across the grain faces.

7. Discussion

Liquid metal embrittlement and cracking of aluminium

alloys by mercury requires the presence of mercury in the

liquid state, tensile stress above a threshold value, which can be

as low as 5% of the yield stress under some conditions for

5083-0 aluminium alloys in mercury, and ‘wetting’ of the

aluminium substrate by the liquid metal, which requires rupture

of oxide films between the substrate and liquid metal. The

fracture surfaces exhibited droplets of mercury throughout, and

these occur after superficial oxidation leads to de-wetting of the

surface. Associated with this is the formation of alumina

whiskers. Both these features when observed on fracture

surfaces provide strong support for LME as the failure

mechanism.

In natural gas processing plants considerable quantities of

mercury can collect in the cryogenic heat exchangers. The

substantial amounts of mercury are derived from traces of

mercury present in natural gas and this can condense and

collect in cold parts of the system if it is not removed in

upstream filters. Typically the mercury would condense onto

surfaces in the solid form (i.e. at temperatures less thanK39 8C) and would only melt during shut down periods when

it would be expected to collect in low points in the manifolds

and pipework in the heat exchanger system.

The key issue for LME is the simultaneous breaching of the

protective alumina layer in the presence of liquid mercury.

Stresses due to internal pressure alone would generally be

above the threshold values for LME. In addition, there are

significant bolting stresses near the flange of the distribution

manifold and residual stresses from welding, resulting in

stresses that could approach the yield stress in this area. Weld

material (with equiaxed grains) is more susceptible to LME

than the pipe material (with elongated grains and a less-favourable intergranular crack path for the relevant crack-plane

orientation). Thus, the high stresses near the weld, more

susceptible weld material, and the potential for mercury to pool

in this region if it is a low spot indicates that LME crack

initiation could occur at the weld.

Rupture of the protective aluminium oxide film, allowing

intimate contact between mercury and aluminium, is most

likely to be caused by slip-bands intersecting the surface due to

micro-yielding and creep processes. This might occur under the

high, sustained pressure and residual stresses but in most

instances is probably accentuated by superimposed thermal-

stress from transient operations and potentially from piping

induced loads.

At high stresses, the oxide-rupture process can sometimes

occur rapidly (!1 h) and can sometimes take considerable

time (O100 h) under nominally the same conditions, for

reasons that are not well understood.

Once cracks have initiated, crack growth in 5083-0 in liquid

mercury can occur very rapidly (up to tens of millimetres per

second)—the time to failure is therefore determined by the

time for cracking to initiate. The variability in crack initiation

times can explain why in some LNG processing plant one

manifold might be extensively cracked, whereas a parallel

manifold might not exhibit any cracking (despite containing

mercury and being subject to similar stresses).It is the largely varying times to crack initiation and the

potentially rapid growth of these cracks once initiated, that

makes the non-destructive inspection for Hg induced LME

problematical. For example, ultrasonic crack detection requires

the plant to be off-line and the cladding removed, and only then

can find cracks once initiated and before they have grown to a

critical size. Radiography can locate areas where Hg is

accumulating providing an opportunity to locally remove the

Hg or to inspect these regions, but while radiography will

clearly show cracks infiltrated with Hg, where such infiltration

occurs, failure is imminent. Hence the non-destructive

inspection for Hg induced LME is quite difficult, suggesting

Fig. 8. Longitudinal fracture surface at the top of one of the ridges showing

lamellar cracking and a more detailed views of the substructure of the fracture

showing additional cracking across the lamellar cracking and tearing with some

boundaries visible and possible sites where second phase particles have been

removed.




that the less well known techniques such as acoustic emission

may be worthy of further investigation for such plant.

The extent of cracking is known to depend, inter alia, on the

amount of mercury present, and cracks can ‘run out of’

mercury and arrest when only small quantities are present.

Large quantities of mercury known to be present in some gas

fields increase the likelihood of failures if there is no mercuryremoval process upstream of the cryogenic heat exchangers.

Most reported instances of mercury induced cracking have

apparently involved relatively small amounts of cracking,

resulting in leaks rather than bursts, probably because there

were only small amounts of mercury present.

Modern LNG plant incorporates various mercury removal

systems, upstream from the cold boxes of the cryogenic

distillation plant to minimise the risk of LME. Cold box

manifolds can be manufactured from aluminium alloys and

tempers that are less susceptible to LME.

8. Conclusion

Liquid mercury in contact with aluminium can induce

serious and rapid failures. Mercury at relatively low

concentrations in natural gas can concentrate in the cryogenic

distillation process. If mercury is present in the liquid form

(e.g. during process interruptions or plant shut down) it can

cause catastrophic damage in aluminium heat exchangers in

LNG plant. The cracking can occur without warning, and

because of the rapid rates of LME crack growth there is no

adequate non-destructive testing techniques to safely monitor

and protect plant.

Mercury induced LME in aluminium alloys leads to

intergranular cracking and potentially very widespread damagein heat exchange manifolds. Cracking surfaces are character-

istically covered with droplets of mercury and in most cases

alumina whiskers are found to grow from cracks.

Acknowledgements

The author wishes to thank Mr Stan Lynch from DSTO for

his assistance in the preparation of this paper.

References

[1] Kamdar MH, editor. Embrittlement by liquid and solid metals. The

Metallurgical Society of AIME; 1984.

[2] Rostoker W, McCaughey JM, Markus H. Embrittlement by liquid metals.

New York, NY: Reihold Publishing Corp.; 1960.

[3] Rhines FN, Alexander JA, Barclay WF. The mechanism of mercury

stress-crack propagation in 70/30 brass and 2024-ST4 aluminum. Trans

ASM 1962;55:23–44.

[4] Nichols H, Rostoker W. On the mechanism of crack initiation in

embrittlement by liquid metals. Acta Metall 1961;9:504–9.

[5] Leeper JE. Mercury—LNG’s problem. Hydrocarbon Process 1980;

59(11):237–40.

[6] Wilhelm SM, McArthur A, Kane RD. Methods to combat liquid metalembrittlement in cryogenic aluminum heat exchangers, Proceedings of

the 73rd GPA annual convention; March 1994. p. 62–71.

[7] Lund DL. Causes and remedies for mercury exposure to aluminum cold

boxes. Proceedings of the 75th GPA annual convention; 1996. p. 282–7.

[8] Nelson DR. Mercury attack of brazed aluminum, heat exchangers in

cryogenic gas service. Proceedings of the 73rd GPA annual convention;

March 1994. p. 178–83.

[9] Gordon, W. Assessing the susceptibility of an LNG plant to mercury-

induced attack. 14th international corrosion congress proceedings, Cape

Town, South Africa; October 1999.

[10] Pinnel MR, Bennet JE. Voluminous oxidation of aluminium by

continuous dissolution in a wetting mercury film. J Mater Sci 1972;7:

1016–26.

[11] Bennet JE, Pinnel MR. Reactions between mercury-wetted aluminium

and liquid water. J Mater Sci 1973;8:1189–93.[12] Bruce L, Wise G. Comments on—voluminous oxidation of aluminium by

continuous distribution in a wetting mercury film. J Mater Sci 1974;9:335.

[13] Phannenstiel L, Phannenstiel L, McKinley L, Sorenson J. ‘Mercury in

natural gas’, Progress in refrigeration science and technology. Proceed-

ings of the 14th international congress of refrigeration, Moscow; 1975.

[14] Fernandes PJL, Clegg RE, Jones DRH. Failure by liquid metal induced

embrittlement. Eng Fail Anal 1994;1(1):51–63.

[15] English JJ, Duquette DJ. Mercury liquid embrittlement failure of 5083-0

aluminum alloy piping. Handbook of case histories in failure analysis,

vol. 2; 1993. p. 207–13.

[16] Lynch SP. Metal-induced embrittlement of materials. Materials charac-

terisation, vol. 28; 1992. p. 279–89.

[17] Gordon P. Metal-induced embrittlement of metals—an evaluation of

embrittler transport mechanisms. Metall Trans A 1978;9A:267–73.[18] Speidel MO. The theory of stress corrosion cracking, NATO,1971.

Feeney JA, Blackburn MJ, editors. Proceedings of the 73rd GPA Annual

Convention, New Orleans; March 1994. p. 389.

[19] Hoagland RG, Liu Y, Benson B. Influence of microstructure on liquid

metal embrittlement of aluminum alloys. ECF 8 fracture behaviour and

design of materials and structures, Torino, Italy; October 1990. p. 316–21.

[20] McIntyre DR, Oldfield JW. Environmental attack of ethylene plant alloys

by mercury, Proceedings of conference ‘corrosion prevention in the

process industries.what European industry is doing’, Amsterdam, The

Netherlands, NACE; 1989. p. 239–52.

[21] Krupowicz JJ, Hampton DS. Technical note: cracking of aluminum alloy

5083 in mercuric salt solutions. Corrosion 1990; 46(2): 159–161.

[22] Kapp JA, Duquette D, KamdarMH. Crack growth behaviourof aluminum

alloys tested in liquid mercury. J Eng Mater Technol 1996;108:37–43.

[23] Turissini RL, Bruno TV, Dahlberg EP, Setterlund RB. Mercury liquidmetal embrittlement causes aluminum plate heat exchanger failure. Mater

Performance; June 1998. p. 59–60.


Corrosión en Plantas de GN

Documents

Transcript of Corrosión en Plantas de GN