acomAVESTA CORROSION MANAGMENT

Evaluation of Materials forSea Water Handling Systems

Continued developments in areas such as offshoreindustry and power generation have led to increaseddemands on materials for sea water handling systems:higher strength, lower weight and better resistance tohigh flow velocities, besides a high corrosion resistance.Systematic investigations of various materials havebeen carried out using different test methods. Theresults from both long term and accelerated tests tendto focus interest on the newer high alloy stainless steelsfor such sea water applications.

The reports in this issue describe different corrosiontest methods carried out in both natural and syntheticsea waters, with concluding remarks regarding choiceof materials. The papers were presented at the NCC/SMOZ workshop "Corrosion Protection of Materials inSea Water Applications" in Amsterdam, November 6-7,1986:

"Sea Water Handling Systems: Past, Present andFuture" by P Gallager, R E Malpas and E B ShonePages 2-5

"An Accelerated Test Method for Crevice Corrosion"by J M KrougmanPages 6-11

"Selection of High-alloyed Steels for Seawater-cooledCondensers" by L M Butter, A H M Keller, H B J KleinAvink and W M M HuijbregtsPages 12-14

All rights reserved. Comments and correspondence can bedirected to Sten von Matérn, Technical Editor, Avesta AB,S-774 01 Avesta, Sweden. Tel. +46(0)226-818 00.Telex 40976 AVESTA S, telefax +46(0)226-545 07. No 3-4 1986

2

acom No.3-4-86

Sea Water Handling Systems:Past, Present and Future

byP. Gallagher, R.E. Malpas and E.B. Shone,

Shell Research Ltd, Thornton Research Centre, PO Box 1, Chester CH1 3SH, UK

Prior to the 1960's galvanized steel was the most com-monly used material in the construction of sea waterhandling systems on ships. However, the cost of re-newing such systems at least twice during the 21 yearlife of a vessel made the more enlightened companiesseriously consider the possibility of using alternativematerials. Economic comparisons were made of pipingsystems fabricated from galvanized mild steel, alu-minium brass and cupro-nickel and this, coupled withthe need to avoid inservice failures led to the progress-ive adoption of more corrosion resistant materials.During this period materials and systems performancewas under continuous review and Table 1 shows theway in which material selection changed in one com-pany (1) between 1950 and 1967.

The 1960's was a very exciting time for marine engin-eers in that larger tankers were being built, 56-250,000dtw and these required bigger engines with largercondenser units and increased flow-through rates ofcooling water. Charter costs were high and any failuresthat either necessitated stops at sea or lengthy drydockings were very costly in terms of lost charter time. Itbecame apparent that failures in the sea water handlingsystems were not acceptable and improved materialsof construction and well designed systems were re-quired. Whilst improvements had been made in thechoice of materials during the 50's and 60's neitherthese nor the designs of the system were adequate forthe vessels currently being constructed. Failures wereoccuring because materials were either being usedbeyond their limitations or unsuitable combinations ofalloys were being used. Several papers have beenpresented (2, 3, 4) that illustrated the problems thatwere being experienced and showed how the damagecould either be avoided or alleviated.

To summarize, it appeared that aluminium brass heatexchanger tubes and pipes were being damaged byerosion/corrosion (impingement attack) and that per-foration was occurring in less than 1 year. It is wellestablished that if sea water velocities exceeded about8 ft/sec that this form of damage occurs; thus the ma-terial was being used beyond its limitation. Obviouslythe situation was not as simple as that, in that other fac-tors had to be taken into account, the most important ofthese being the removal of ferrous material from the sea

water handling systems. This resulted in the eliminationof the beneficial effects that ferrous components have(cathodic protection and protective film formation) onthe performance of aluminium brass. Once theseeffects were fully realized iron was again introduced intothe system in the form of sacrificial soft iron anodes inwater boxes and sacrificial spool pieces in pipelines.Additionally ferrous sulphate solution was added to thecooling water to allow protective films to form on thealuminium brass surface. These measures were verysuccessful in that failures ceased almost immediatelythey were applied, thus the introduction of iron intothe system had raised the threshold water velocity atwhich impingent damage to aluminium brass occurred.Further work still requires to be carried out to quantifythe benefits of these remedial measures and establishthe new limits at which aluminium brass can be usedwithout the possibility of failure by impingement attack.

As manpower was at a premium on board ships, ferroussulphate dosing was not always popular with operatorswho were continually striving to further reduce manninglevels. Additionally once the waterside problems hadbeen solved and the useful life of the aluminium brasshad been extended, newer steamside problems (5)associated with hydrazine attack and impingementbecame apparent. It was then obvious that alternativematerials of construction were required if these prob-lems were to be overcome. In the early 70's trials werecarried out using 90/10 copper/nickel (BS 2871 Part 3CN 102) pipework and 70/30/2/2 copper/nickel/iron/manganese (BS 2871 Part 3 CN 108) heat exchangertubes. After early problems associated with the qualityof tubes in the as supplied condition and precommis-sioning treatments had been optimized, the perform-ance of pipes and heat exchangers constructed in thisway was found to be excellent with only 149 tubesrequiring plugging out of 161,000 over a 10 year periodand pipework failures being even more unusual.

In a similar way the problems encounted during the1960's with gunmetal and cast iron valves were in themain overcome by ensuring that cast iron valves werenot connected directly to copper alloy pipelines thusavoiding bimetallic corrosion. Similarly in some caststeel valves the discs were trimmed with stainless steelweld deposit and after only a few months in service

Table 1: The progressive adoption of more corrosion resistant materials in one company during the period 1950-1967

Period of construction Pipe material Valve and pump material Bend material Heat exchanger material

1950-54 Copper Gunmetal Cast iron Aluminium brass1952-55 Aluminium brass Gunmetal Cast iron Aluminium brass

95/5 Cu.Ni.Fe.1955-60 95/5 Cu.Ni.Fe. Gunmetal Cast iron Aluminium brass1962-63 95/5 Cu.Ni.Fe. Gunmetal Gunmetal Aluminium brass1963 90/10 Cu.Ni.Fe. Gunmetal Gunmetal Cu.Ni.Fe. Aluminium brass1966-67 90/10 Cu.Ni.Fe. Gunmetal Gunmetal Cu.Ni.Fe. Aluminium brass

3

acom No.3-4-86

severe corrosion of the cast steel occurred. Failures ofgunmetal globe valves occurred because of erosion-corrosion possibly intensified by throttling of the valve inservice which would have increased turbulence in thevalve.

These problems and others were in general overcomeby recommending that valves must be lined with eithera suitable rubber or plastic and that the disc in butterflyvalves must be made from an alloy with at least the cor-rosion resistance of a good quality nickel aluminiumbronze. Valves of this type offered an economical sol-ution to the problem.

The gunmetal pump bodies usually specified werebeing severely damaged by erosion/corrosion as werethe impellers. Whilst some manufacturers tended to at-tribute such damage to the design of the systems andthe higher speeds of vessels, the more enlightenedagreed that the failures were associated with the use ofmaterials whose properties and in some instancesquality, were inadequate for their intended application.By ensuring that only sound castings were used it waspossible to extend the useful lives of pumps to anacceptable level but the problem was not fully over-come and the need to change impellers on a regularbasis still exists.

By the mid 70's the selection of materials for use in seawater handling systems on ships had reached a state ofdevelopment where in general their performance wasregarded as being satisfactory.

The 70's saw the beginning of the North Sea oil produc-tion. Unfortunately speed was of the essence in that itwas essential to ensure that the platforms began pro-ducing on time. This meant that the technology that hadbeen developed for shipboard applications could notalways be used since materials were often not availablein the sizes and quantities required. Alternatives had tobe used, for example copper nickel piping had to besubstituted with plain carbon steel piping both with andwithout linings. Its performance was no better thanresearch and experience could predict, in that rapidcorrosion occurred on the uncoated pipes and atdefects in the plastic or paint lined pipes. Other prob-lems of a similar nature and some more unusual onesthat occurred are described elsewhere (6). By the late70's the selection of materials for offshore applicationswere well established and the copper alloy systems onplatforms were performing satisfactorily. In generalterms the specification asks that all pipework should bein 90/10 copper-nickel with heat exchanger tubes ofeither 90/10 or 70/30 copper-nickel, the exact choice ofalloy depending upon the conditions prevailing withinthe system.

Whilst the current requirements of both the oil shippingindustry and the North Sea oil production industry aremet by the copper alloys it is very doubtful that they willbe adequate for the future. It seems very likely that asgas and oil production extends into deeper waters thenit will become increasingly important to reduce theweight and size of topside facilities on offshore plat-forms. Similarly light weight, compact systems with seawater flowing through them at high speeds could bedesirable in the shipping industry. The inherent prop-erties of copper alloys will probably exclude them fromuse in sea water handling systems of the future sincethey are prone to erosion damage at sea water veloc-ities exceeding about 3.5 m/s and have strength limita-tions, which prohibits their usage in thin sections.

Several previous papers (7, 8, 9,10,11) have indicatedthat the "newer" high alloy stainless steels appear to be

attractive alternatives to the currently used copperalloys and our work continues to indicate that the futurelies with these alloys. A paper by Eriksen (12) illustratesthe weight and cost savings that can be made by using astainless steel of the 254 SMO type instead of the moretraditional 90/10 copper-nickel. We have carried out asimilar exercise and if we assume that because of thesuperior corrosion resistance and strength a 20" cop-per-nickel schedule 40 pipe can be replaced by a 14"pipe in a stainless steel containing about 20 %w Cr,19%w Ni, 6%Mo, 0.2 %N (254 SMO type) then the costsof a 50 m length would be as shown in Table 2. From theeconomics point of view there appears to be little doubtthat stainless steels are viable alternatives to copperalloys for the construction of sea water handling sys-tems. However, we believe that several questions stillhave to be answered before stainless steel(s) can beused with confidence as alternatives to copper alloys.We have attempted to answer some of these and in thenext few minutes intend to consider some of the ques-tions and provide some of the answers.

Perhaps the most important question is, which stainlesssteel(s) can be recommended and to what composi-tions) should it (they) conform.

Table 2: Cost comparison

Cost of 50 M lengths of schedule 40 pipe

20" 90/10 Copper nickel alloy £1200/m14" 254 SMO £ 793/m14" Alloy of similar composition to 254 SMO £ 658/m

To help answer this question over 50 alloys have beenevaluated to establish their corrosion characteristicsand in particular their ability to resist crevice corrosion.All of this work has been carried out in natural flowingsea water since very early work showed that testscarried out in artificial sea waters produced erroneousresults, whilst those undertaken in natural sea waterthat had been transported must be treated with cautionin that they may be misleading.

This work is showing that the susceptibility of severalstainless steel alloys to crevice corrosion is low and thatone of the most important alloying elements is nitrogenprobably in combination with molybdenum. From Table3 (page 4) the synergistic influence of molybdenum andnitrogen can be seen. Further work is required beforethe corrosion characteristics can be confidently pre-dicted by reference to the analysis of the alloy.

Since castings, forgings, wrought and welded materialwill be required (valves, pumps, flanges, pipework andheat exchangers) in a sea water handling system it isessential to establish the corrosion characteristics ofstainless steels manufactured by these various routes.We believe that the most effective method of carryingout such work is to evaluate pumps, valves, etc. in proto-type sea water handling systems. These are currentlybeing evaluated at our test site in Holyhead and theresults of this work will be reported in the future.

It is inevitable that within a sea water system more thanone alloy type will be chosen because of varyingstrength requirements or even availability in a particularform. Electrical isolation of sections of the system can-not be relied upon offshore since electrical safety re-quirements (earthing) predominate on a steel structure.It is therefore essential that materials selected for thesea water system are galvanically compatible.

4

acom No.3-4-86

Table 3: Crevice depth and area of attack onstainless steel alloys

With this in mind we have carried out a research pro-gramme which showed that:

a) the materials which are intrinsically not susceptibleto crevice corrosion when exposed to sea water andnot coupled to another alloy are also resistant to thisform of corrosion when coupled even to more noblematerials.

b) For alloys which are intrinsically susceptible tocrevice corrosion the propagation rate of this form ofattack can be increased by coupling noble alloys tothem.

Details of this work were presented elsewhere (13) butin essence it means that if we know and understand theintrinsic corrosion characteristics of an alloy then wecan design a system using galvanically compatiblematerials.

Sea water systems are chlorinated to prevent foulingwhich if not checked can cause blockage in a very shorttime. On offshore structures a residual chlorine level ofbetween 0.5 and 1.0 ppm is maintained within the sys-tem using electrochlorinators. Chlorine is introducedinto the caisson below each submersible feed pumpwith suitable failsafe valves to prevent a build up ofchlorine in the event of a pump failure. With copper alloysystems provided the chlorine levels are maintained inthe range stated above no chlorine related problemshave been encountered. Some work on the influence ofchlorination on the corrosion characteristics of selectedstainless steels has already been reported (14). Theseresults are very encouraging in that they indicated thatalloys, such as 254 SMO, with a low pitting probability athigh potentials are even less likely to pit since theequilibrium rest potential is lowered by chlorination.Materials such as AISI 316L, with a high pitting prob-ability at low potentials remain very vulnerable. How-ever, between these two extremes of alloys we have agroup of materials such as certain duplex steels whichshow high pitting probabilities at high potentials and asignificantly lower pitting probability as the potential isreduced. Since chlorination reduces the potential it willreduce the possibility of pitting initiation. Chlorination atlow residual levels should thus be beneficial to the cor-rosion performance of alloys of this type.

The reason for the drop in potential on chlorination isuncertain; however, one likely explanation is a change(a decrease) in the kinetics of the cathodic (oxygenreduction) reaction accompanying the death of marineorganisms on chlorination. It has been suggested pre-viously (15) that these organisms can catalyse oxygenreduction possibly by enzyme action, resulting in a rela-tively positive mixed rest potential in the system. It isinteresting to note that the potentials observed in thesechlorinated systems are very similar to those seen inartificial sea water where marine organisms are alsoabsent. We have observed many times in the past thatcorrosion in artificial sea water is considerably less thanthat seen in natural sea water, lending support to ourideas about the advantage of chlorination.

Finally, it should be recognized that these possibleadvantages of chlorination will only apply at low residuallevels. As chlorine concentrations rise and pass thelevel of oxygenation in the system (approximately 8ppm), chlorine reduction will become the favouredcathodic reaction and will be able to support a higheranodic current. The potential of the system will thustend to rise and corrosion will increase. At very highchlorine levels the extent of corrosion could of coursebe very severe. The general potential-chlorine relation-ship should thus resemble Figure 1.

Figure 1The influence of chlorine additions on the rest potential ofstainless steel.

Another very interesting aspect of our recent work is theinfluence that temperature has on the time to initiationof crevice corrosion of stainless steels. Figure 2 showsthat the time to initiation of an AISI 316 austenitic or2205 type duplex is considerably reduced as the tem-perature of the sea water increases. Whilst the reasonfor this is not fully understood and effects such as oxy-gen content of the sea water, degree and type of foulinghave to be considered, it is a little alarming in that it mayindicate that stainless steels that appear to be resistantto corrosion at up to 10°C may initiate pits at high tem-perature. In the Northern North Sea between 60°N and62°N the mean sea surface temperatures recordedbetween 1973 and 1982 were found to vary from a mini-mum of 8.0°C in January/March to a maximum of 13.0°Cin August. Monthly minimum, mean and maximumvalues are given in Table 4. This suggests that testscarried out in sea water at about 15°C should provideresults sufficiently realistic to simulate North Seaconditions. However, these results should not beregarded as meaningful for hot climates such as theChina Sea where sea water temperatures may be up to30°C.

5

acom No.3-4-86

References

Figure 2.Effect of temperature on the time to initiation of selectedstainless steels.

Table 4: Northern North Sea- sea surface temperature

Month Minimum Mean Maximum

Jan 6.5 8.0 9.0Feb 7.0 8.5 9.0Mar 6.0 8.0 9.0Apr 7.0 8.0 9.5May 7.0 9.0 11.5Jun 6.5 10.0 12.5Jul 9.5 12.0 14.0Aug 12.0 13.0 15.0Sep 10.0 12.0 13.5Oct 5.5 10.0 11.5Nov 7.5 9.5 10.5Dec 6.0 8.5 9.5

In this short introductory paper we have tried to presentthe current state of the art for sea water handling sys-tems and indicate where the future lies. Additionally wehave highlighted some of the problems associated withthe newly developed stainless steels and indicated thatseveral questions have to be answered before they canbe used without reservation. However, if appropriateresearch is undertaken in a realistic and meaningful waythen the results will be provided that will ensure that inthe future we will be able to utilise stainless steels andhave reliable light weight sea water systems.

1 W.H. Falconer and LK. Wong, Sea water systemsInst. Mar. Eng. Materials section symposium, March1968.

2 P.T. Gilbert and W. North, Copper alloys in marineapplications Trans. Inst. Marine. Eng. 1972, Vol. 84.

3 R.A. Connel and E.B. Shone, Sea water circulatingsystems. Supplement to chemistry and industry,2 July 1977.

4 B. Todd and P.A. Lorett, Selecting materials for seawater systems, Marine Engineering practice, Vol. 1,Part 10, Inst. Mar. Eng.

5 E.B. Shone and G.G. Grim, 25 years experience withsea water cooled heat-transfer equipment in theShell Fleets.

6 E.B. Shone, Salt water systems for offshore plat-forms. EFC Conference, Nice, November 1985.

7 H.P. Hack, Crevice corrosion behaviour of 45 molyb-denum-containing stainless steels in sea water, Cor-rosion 82, NACE, Houston, Texas, March 1982.

8 M.A. Streicher, Analysis of crevice corrosion datafrom two sea water exposure tests on stainlesssteel alloys, Materials Performance, May 1983.

9 A.I. Asphahani et al, Highly alloyed stainless ma-terials for sea water applications, Corrosion 80NACE, Chicago, Illinois, March 1980.

10 A.P. Bond and H.J. Dundas, Stainless steels for seawater service, Stainless steels 77.

11 A.J. Sedriks, Corrosion resistance of austenitic Fe-Cr-Ni-Mo alloys in marine environments, Interna-tional Metals Review.

12 T. Eriksen, Weight optimisation in offshore construc-tion, Corrosion control in offshore environments,Stavanger, August 1984.

13 E.B. Shone and P. Gallagher, Galvanic compatibilityof selected high alloy stainless steels in sea water.Seminar on High Alloy Stainless Steels for CriticalSea Water Applications. Chemical Industry. Birming-ham, March 1985.

14 R.E. Malpas, P. Gallagher and E.B. Shone, Corrosionand chlorination in materials for offshore systems.Conference on chlorination of sea water systemsand its effects on corrosion. Society of ChemicalIndustry, Birmingham, March 1986.

15 V. Scotto, R. Di. Cinitis and C. Marcenero, Theinfluence of marine aerobic microbial film on stain-less steel corrosion behaviour. Corrosion Science,25. 185 (1985).

6

acom No.3-4-86

An Accelerated Test Methodfor Crevice Corrosion

byJ.M. Krougman,

Corrosion Laboratory, Royal Netherlands Naval College, 1781 AC Den Helder, The Netherlands

DescriptionThe aim of the development of the test method was toincrease reproducability of crevice corrosion testing.This purpose has been attained by means of creatingconditions of maximum severity and increasing theprobability of reaching critical conditions for crevicecorrosion.

Within given chemical, metallurgical and environmentalconditions, crevice corrosion can be stimulated. Thusfor crevice geometry critical dimensions have to be metwhile the potential to be applied has the highest valuestainless alloys can reach. Temperature may also effectcrevice corrosion, in this respect the crevice corrosiontemperature has been defined as the temperaturebelow which no crevice attack will occur. On this basiscrevice corrosion has been stimulated by anodic polar-ization in sea water. In order to favour crevice corrosioninitiation, the crevice area was provided with a largenumber of scratches (fig. 1) which by covering with arubber O-ring acted as micro crevices (fig. 2).

ResultsTwo groups of iron and nickel base alloys with varyingmolybdenum content were selected for testing(table 1).

In brackish sea water the specimens were polarizedpotentiostatically at 500 mV SCE. The resulting currentflowing between specimen and a platinum electrodewas recorded as a function of time. Moreover the cur-rent was integrated to determine the total charge. Thetests were carried out in fivefold at 10, 25, 40, 55 and70°C. The duration of the tests was three hours. Mostresistant alloys were 254 SMO and INCONEL 625 whichremained free from crevice corrosion up to 40°C (fig. 3).All specimens which initiated suffered from crevice cor-rosion preferentially in the scratched areas (fig. 4). Thereproducability of initiation was convincing becausefrom 76 % of the attacked material and temperaturecombinations at least 3 specimens initiated and from53 % of the combinations all 5 specimens initiated.Propagation data have been indicated by maximumpenetration depths, mass losses and total charges(fig. 5). It was obvious that propagation increased withtemperature. It appeared that propagation decreasedwith molybdenum content.

The initiation temperature of the tested alloys can becompared with the results of other authors. It appearedthat initiation temperatures obtained from the presenttests fit best with those of Garner (table 2).

Figure 1Scratching assembly.

Figure 2Holder for micro creviced specimen.

Table 1: Nominal chemical composition for austeniticstainless steels and nickel alloys tested.

Alloy Element % wt

Cr Ni Mo Cu Fe Other

AISI 316L 18 10 2.5 Bal1713 NCN 17 13 4.5 Bal 0.15 NA 963 17 16 6 1.6 Bal 0.15 N254 SMO 20 18 6 0.7 Bal 0.20 NAL 6X 20 24 6 Bal1925 hMo 21 25 6 1.7 Bal 0.14 NINCOLOY 825 22 Bal 3 2.2 30 1 TiHASTELLOY G 22 Bal 6 1.8 20 2 Cb+Ta;

1.5 CoINCONEL 625 22 Bal 9 2.5 3 Cb+Ta

7

acom No.3-4-86

Figure 3 wwPotentiostatic exposure tests with micro creviced austeniticstainless steels and nickel alloys at 500 mV SCE in brackishsea water.

Figure 4 uuSpecimens of austenitic stainless steels and nickel alloys afterpotentiostatic exposure at 500 mV SCE in brackish sea water.

Figure 5Maximum penetration depths of crevice corrosion, masslosses and total charges of austenitic stainless steels andnickel alloys at 500 mV SCE in brackish sea water.

Table 2: Initiation of crevice corrosion of austeniticstainless steels and nickel alloys in ferriticchloride tests, oxydizing sodium chloride -hydrochloric acid tests and potentiostatictests in brackish sea water.

Material Ferric chloride Oxydizing Brackishsodium sea waterchloride

Garner Nikhil hydro-et al. chloric acid

AISI 316L -5- 0°C ≤10°C1713 NCN 10-15°C 11-25°CA 963 11 -25°CAL 6X 30-35°C 37°C 25°C 11 -25°C1925 hMo 35-40°C 26-40°C254 SMO 40-45°C 46°C 30°C 41-55°CINCOLOY 825 -10- 0°C ≤-5°C ≤10°CHASTELLOY G 15-20°C 25°C ≤10°CINCONEL 625 45-50°C 41-55°C

8

acom No.3-4-86

AdvantagesThe main advantage of this test method is attainmentof increased reproducability of test results. Next to thisthe limited time effort involved and the ability to differ-entiate between alloys which do not initiate in exposuretests are advantageous.

These points are highlighted with the testing of duplexstainless steel casting, bar and plate in brackish seawater. The testing consisted of exposure of micro crev-iced specimens which were either galvanically coupledwith non micro creviced plates of the same alloy (fig. 6)at 25°C or potentiostatically polarized at 500 mV SCE at10, 25 and 40°C.

In the exposure tests with galvanically coupled speci-mens casting initiated within ten days, bar and plateremained free from crevice corrosion initiation (fig. 7and table 3). In the potentiostatic tests casting initiatedat 10°C, bar at 25°C and plate at 40°C (fig. 8 and table 4).

In both types of exposure tests casting appeared to bethe least resistant against crevice corrosion initiation.The potentiostatic tests indicated that plate is moreresistant than bar. So the product form may notablyinfluence the susceptibility to crevice corrosion.

It lasted eight weeks to obtain the data from the expo-sure tests with galvanically coupled specimens. How-ever, the data from the potentiostatic tests werealready available after nine times three hours.

LimitationsA limitation of the potentiostatic tests with micro crev-iced specimens is that the results hold just for the speci-fic test conditions. So the data obtained can be appliedonly for ranking purposes. Although ranking can behelpful in materials selection problems, generally nomore detailed insight into the expected performance increvice conditions is obtained.

Other more experimentally dedicated limitations arethe choice of the test potential and the composition ofthe sea water to be used for testing. In sea water themaximum rest potential depends on temperature(fig.9). So for every test temperature in the potentio-static test the corresponding maximum potentialshould be determined. In artificial sea water the propa-gation rate of crevice corrosion was found to be higherthan in brackish sea water (fig. 10 and table 5). Anattempt was made to explain this difference from thedifference in salinity between artificial sea water andbrackish sea water.

However, from potentiostatic exposure tests in artifi-cial sea water, brackish sea water, diluted brackish seawater and concentrated brackish sea water it appearedthat variations in the salinity of brackish sea waterbetween 17.5 and 35.0% do not influence initiation andpropagation significantly in the potentiostatic test(fig. 11 and table 6) (page 10).

Figure 6Holder for plate to be galvanically coupled with micro crevicedspecimen.

Figure 7Exposure tests with micro creviced duplex stainless steel(25 % Cr, 6 % Ni, 3 % Mo) galvanically coupled with plateof the same alloy in brackish sea water at 25°C.

Table 3: Results of exposure tests with micro crevicedduplex stainless steel (25% Cr, 6°/o Ni, 3% Mo)galvanically coupled with plate of the samealloy in brackish sea water at 25°C.

Product Mass Total Maximum Typeform loss charge penetration of

(mg) (C) depth localized(µm) attack*

Casting 144.7 543.20 700 c, pBar <0.7 0.17 <3 PPlate <0.1 0.15 0

* c = crevice corrosion p = pitting 0 = no localized attack

9

acom No.3-4-86

Figure 8Potentiostatic exposure tests with micro creviced duplexstainless steel (25 % Cr, 6 % Ni, 3 % Mo) at 500 mV SCE inbrackish sea water at 10, 25 and 40°C.

Table 4: Results of potentiostatic exposure tests withmicro creviced duplex stainless steel (25% Cr,6% Ni, 3% Mo) at 500 mV SCE in brackish seawater at 10, 25 and 40°C.

Product Tem- Mass Total Maximum Type ofform perature loss charge penetra- localized

tion depth attack*(°C) (mg) (C) (µm)

Casting 10 0.1 0.36 7 c, p25 0.2 0.66 30 c, p40 21.8 83.30 220 c, p

Bar 10 <0.1 0.01 <3 P25 <0.1 0.07 10 c, p40 1.8 7.42 150 c, p

Plate 10 <0.1 0.01 025 <0.1 0.02 040 <0.1 0.03 10 c

* c = crevice corrosion p = pitting 0 = no localized attack

Figure 9Restpotential as a function of temperature of ferritic stainlesssteel (25 % Cr, 4 % Ni, 4 % Mo) in brackish sea water afterexposure during 30 days at 22°C.

Figure 10Current as a function of time at 500 mV SCE of AISI 316Land 254 SMO in brackish sea water and artificial sea water at25°C.

Table 5: Results of potentiostatic exposure tests in brackish sea water (BS) and artificial sea water (AS) at25°C and 500 mV SCE.

Material Mass loss (mg) Total charge (C) Maximum penetration Type of localizeddepth (µm) attack*

BS AS BS AS BS AS BS AS

AISI 316L 19.4 39.0 74 148 300 410 c, p c, p16.5 35.8 63 136 315 415 c, p c, p21.5 36.7 82 140 360 380 c, p c, p19.1 37.5 73 143 400 390 c, p c, p21.1 32.3 79 123 270 390 c, p c, p

254 SMO <0.1 <0.1 8.8X10-3 9.8X10-3 <1 <1 none none<0.1 <0.1 8.6X10-3 8.9X10-3 <1 <1 none none<0.1 <0.1 9.6X10-3 9.9X10-3 <1 <1 none none<0.1 <0.1 8.0X10-3 8.7X10-3 <1 <1 none none<0.1 <0.1 9.1 X10-3 8.5X10 3 <1 <1 none none

* c = crevice corrosion p = pitting

10

acom No.3-4-86

Figure 11Potentiostatic exposure tests (in duplicate) with micro crev-iced stainless steel (17 % Cr, 16% Ni, 6% Mo) at 500 mV SCE inartificial sea water (AS), brackish sea water (BS), dilutedbrackish sea water (DBS) and concentrated brackish seawater (CBS) at 45°C.

Table 6: Results of potentiostatic exposure tests withmicro creviced stainless steel (17 % Cr,16 % Ni, 6 % Mo) at 500 mV SCE in artificial seawater (AS), brackish sea water (BS), dilutedbrackish sea water (DBS) and concentratedbrackish sea water (CBS) at 35, 40 and 45°C

Environ- Tem- Mass Total Maximum Type ofment perature loss charge penetration localized

(°C) (mg) (C) depth (µm) attack*

AS 35 <0.1 0.05 20 c40 <0.1 0.25 35 c45 0.7 2.61 75 c

BS 35 <0.1 0.03 15 c40 <0.1 0.25 45 c45 0.3 1.04 65 c

DBS 35 <0.1 0.03 10 c40 <0.1 0.19 40 c45 0.3 1.17 65 c

CBS 35 <0.1 0.02 15 c40 <0.1 0.12 40 c45 0.2 0.69 70 c

ApplicationThe final conclusion is that for stainless alloys selectionfor marine service more extended crevice corrosiontesting is required than one single test method. Accord-ingly a test procedure for a practical case might consistof:

1. exposure of micro creviced specimens galvanicallycoupled with non micro creviced specimens of thesame alloy in sea water at, in general, the maximumservice temperature (table 3);

2. potentiostatic exposure of micro creviced speci-mens in sea water at the maximum service tempera-ture and at lower and higher temperatures (table 4);

3. anodic polarization measurements of the alloy insimulated crevice solutions at the maximum servicetemperature and at lower and higher temperatures(fig. 12).

References1 J.W. Oldfield and W.H. Sutton: Br. Corr. J. 13 (1978)

13.2 F.P. Usseling: Br. Corr. J. 15 (1980) 51.3 R.M. Kain: CORROSION, Anaheim (1983) paper no.

203.4 J.M. Krougman and F.P. IJsseling: Proc. 6th Intern.

Congr. on Marine Corros. and Fouling, Athens (1984)75.

5 T.S. Lee and K.L Money: Mat. Perf. 8 (1984) 28.6 A. Garner: NACE Corros. Conf., Houston (1982)

paper no. 195.7 R.M. Kain: Mat. Perf. 2 (1984) 24.8 R.M. Kain, A.H. Tuthill and E.C. Hoxie: J. Mat. for

Ener. Syst. 4 (1984) 205.9 R.M. Kain, T.S. Lee and J.W. Oldfield: CORROSION,

Boston (1985) paper no. 60.10 M.D. Carpenter, R. Francis, L.M. Phillips and J.W.

Oldfield: Br. Corr. J. 21 (1986) 45.11 T.S. Lee, R.M. Kain and J.W. Oldfield: Mat. Perf. 7

(1984) 9.12 J.M. Krougman and F.P. IJsseling : Proc. 137th Event

of EFC, Ferrara (1985) 135.

* c = crevice corrosion p = pitting

11

acom No.3-4-86

Figure 12Polarization measurements with duplex stainless steel (25%Cr, 6 % Ni, 3 % Mo) in simulated crevice solutions at 25°Cscanning rate 20 mV min-1; top artificial sea water pH 8; bot-tom artificial seawater+ 146 g.l-1 NaCI+ HCI, pH 2.

12

acom No.3-4-86

Selection ofHigh-alloyed Steels for

Seawater-cooled Condensorsby

L.M. Butter, A.H.M. Keller, H.B.J. Klein Avink, W.M.M. Huijbregts,N.V. KEMA, Research and Development Division, Department of Chemical Research,

P.O. Box 9035, 6800 ET Arnhem, The Netherlands

High-alloyed steels can be applied in seawater-cooledcondensers, if they have great resistance to pitting andcrevice corrosion. To obtain appropriate selection cri-teria the results of electrochemical measurementshave been compared with those of exposure tests inartificial and natural seawater. A great number of steelshave been tested. See the table on page 14, in whichthe steels and the results are summarized. For somesteels more heats were tested.

ElectrochemicalCharacterizationPitting Corrosion

The resistance to pitting corrosion has been deter-mined by means of polarization curves, measured inartificial seawater at a low potential scanning rate(3 mV/min).

On the basis of their pitting potential values the steelshave been divided in three categories:- pitting potential lower than 900 mV+ pitting potential higher than 900 mV, but lower than

1200mV++ pitting potential higher than 1200 mV

The pitting potentials of the AL6X, 29-4C and 254 SMOsteels are high (>1200 mV). Most heats of Monit 2502also have high pitting potentials, but some show valuesin between 900 and 1200 mV.

Crevice CorrosionCondenser tubes can be blocked by spongy rubberballs, pocks, eels, etc. This may result in crevice forma-tion inside the tube. Hydrolysis of corrosion productswill result in a pH-decrease in the crevice. Even high-alloyed steels can lose their passivity in this acid envi-ronment. The active-passive behaviour of the steels hasbeen determined by means of polarization curves,measured in acidified 6 % Nad-solutions.

The acid concentrations in which the steel remainedpassive and became active are shown in the followingfigure.

13

acom No.3-4-86

Both heats of 254 SMO were passive in very high acidconcentrations. One of the heats was still passive evenin 1.25 Mol HCI. The materials Sanicro28 and Monit 2502showed more variations in critical acid concentrations.See also the table on page 14 (All mV-measurementsare made against NHA).

Exposure in SeawaterCrevice-type tube samples were exposed in artificialand in natural seawater. In case of a blocked condensortube, the seawater temperature in the crevices wouldbecome equal to that of the condensed steam (40°C).The seawater in our experiments was heated up to thistemperature.

Artificial SeawaterThe samples from the laboratory exposure tests havebeen inspected frequently. The larger part of the ma-terials showed corrosion in some degree. Only 254SMO and AL6X were not corroded, although they hadbeen exposed during 75 weeks. See table for results.

Natural SeawaterFor the exposure in natural seawater 6 sites werechosen along the Dutch coast.

The samples were examined for corrosion after 1.5years of exposure. The corroded surface areas weremeasured and plotted in the diagram at the bottom ofthis page.

Judging from the corroded area of the samples the cor-rosivity of the water on the 6 sites decreased as follows:

Terneuzen → Borssele → Eemshaven →Maasvlakte → Texel → IJmuiden

The differences were rather small. On the IJmuiden sitethe harbour water was used.

Samples with only very small corroded areas wereexamined in the Scanning Electron Microscope (SEM).The corroded spots had an etched surface. Only a smalldefect on the 254 SMO material had no etched surface.It rather seemed to be mechanical damage. The depthprofiles on the corroded spots were measured bymeans of the SEM technique.

14

acom No.3-4-86

Trade Heat Electro- Exposure

Name chemistry in Seawater

1 2 3 4 5

x2CrNi 18 9 a - --

x5CrNiMo 18 10 5R60 a - - - --

x2CrNiMo 18 10 3R60 a - - --

x3CrNiMoN 17 13 5 A400 a - - --

x2NiCrMoCu 25 20 5 UR. B6 a - -

2RK65 a - - - --

b + + - --

A962 a - -

b - -

c - -

x2CrNiMoN 22 5 UR. 47 a - - --

x4CrNiMoCu 17 16 6 A963 a - - -

x1NiCrMoCu 31 244 San 28 a - - - - 145

b + + + + - 10

x2CrNiMo 25 4 4 Monit a + + + + + +

2502 b + + - - +

c + + - + - 50

d + + - + - 30

e + - - +

f + - - 50

g + + - +

x2CrNiMoCu 20 18 6 254 SMO a + + + + + - 40

b + + + + + +

x2CrMo 29 4 29-4C a + + + + - 25

b + + + + -

c + + - + +

d + + - + +

e + + - - - 50

x2NiCrMo 24 20 6 AL6X a + + + + +

Titanium a + + + + +

Conclusions

1 Pitting potential <900 mV: -; >900 mV: +; >1200 mV: + +2 Active behaviour in HCI Mol: <0.75: -; ≥ 0.75: +; >1.0: + +3 Exposure in artificial seawater after 75 weeks

corrosion = -; no corrosion = +Exposure in Natural Seawater4 Corroded surface area: 0 mm2: +; 0-10 mm2: -; >10 mm2: --5 Depth of corrosion spots (µm)

The two heats of the 254 SMO steel have the greatestcorrosion resistance.

The AL6X steel also has great corrosion resistance.Only one heat of this material was tested. The 254 SMOsteel is considered a good alternative material forTitanium condensor tubes.

The Sanicro28, Monit2502 and 29-4C steels showedrather high variations in corrosion resistance. Pre-selec-tion of the material by means of laboratory tests, suchas electrochemical measurements, is necessary.

Electrochemical measurement of pitting potential andcritical acid concentration for passive behaviour is con-sidered to be a good selection test. For the applicationof stainless steel in seawater-cooled condensers apitting potential of at least 1200 mV and a critical acidconcentration for passive behaviour of 1 Molar are pro-posed as selection criteria.

Poster presentation at the NCC/SMOZ-conference

15

acom No.3-4-86

acom is distributed free of charge to personsactively involved in the development of the processingindustry and other areas where stainless steels are im-portant.

to Sten von Matérn, Avesta AB, S-774 01 Avesta, Sweden

acom appears four times a year, and we welcomeyour application, as well as additional applicationsfrom your friends and colleagues.

Name: ___________________________________________________________ Title: ____________________________________

Company: ________________________________________________________ Position:__________________________________

Mailing address: ______________________________________________________________________________________________

Postcode/City: _______________________________________________________________________________________________

Country: ____________________________________________________________________________________________________

£ Please, add my name to your address list acom No. 3-4-86