international journal of refrigeration 31 (2008) 742–747

available at www.sciencedirect.com

w w w . i i fi i r . o r g

journal homepage: www.elsevier.com/locate/ijrefrig

Review

Stress corrosion cracking in refrigeration systems

Andy Pearson*
Star Refrigeration Ltd., G46 8JW Glasgow, UK

article info abstract

Article history: It is generally understood that stress corrosion cracking can occasionally affect the high
Received 17 June 2007 pressure vessels in ammonia refrigeration systems, but the nature of the problem is often
Received in revised form mis-represented in design codes, safety standards and textbooks and ways in which the
26 November 2007 initiation of stress corrosion can be prevented are ignored. It is also not appreciated that
Accepted 28 November 2007 stress corrosion can also affect low pressure vessels and pipework in ammonia systems,
Published online 5 December 2007 and copper pipework in fluorocarbon refrigeration systems.
This paper provides an overview of published literature on stress corrosion cracking and
Keywords: then presents some recent case study material illustrating various ways in which stress
Refrigeration system corrosion has arisen in practice. The conclusions will provide guidance for design engi-
Ammonia neers and equipment owners on eliminating stress corrosion cracking and on dealing
Survey with it if it arises.
Crack ª 2007 Elsevier Ltd and IIR. All rights reserved.
Corrosion
Heat exchanger
Pressure

Fissures dues à la corrosion dans les systèmes frigorifiques

Mots clés : Système frigorifique ; Ammoniac ; Enquête ; Fissure ; Corrosion ; Échangeur de chaleur ; Pression

1. Introduction ammonia vessels operate in corrosive conditions and under

moderate to high stresses without developing stress corrosion
Stress corrosion cracking is defined by Cottis (2000) as ‘‘crack- cracks. The terms ‘‘environmentally-assisted fracture’’ or ‘‘envi-
ing due to a process involving conjoint corrosion and straining of ronmentally-assisted cracking’’ are used to cover a range of pro-
a metal due to residual or applied stresses.’’ He emphasises that cesses, including liquid metal embrittlement, liquid metal
‘‘straining’’ should not be assumed to indicate plastic deforma- assisted cracking, stress corrosion cracking, corrosion fatigue
tion as SCC is often found under elastic strain conditions with (cracking under the influence of fluctuating stress and corro-
no indication that the material has ever reached the elastic sion), and a range of fretting and wear processes. Cottis iden-
limit. This definition is not very helpful because many tifies three factors required to produce stress corrosion

* Tel.: þ44 141 638 7916; fax: þ44 141 638 8111.
E-mail address: [email protected]
0140-7007/$ – see front matter ª 2007 Elsevier Ltd and IIR. All rights reserved.
doi:10.1016/j.ijrefrig.2007.11.015
 international journal of refrigeration 31 (2008) 742–747 743

cracking: susceptible material, appropriate environment and overstressing or locked-in stresses.’’ Dawson noted that of 58
sufficient tensile stress. failed vessels, 85% had stress corrosion cracks in the end
caps and 10% had stress corrosion cracks in the shell. In the
remainder of cases the crack location was not identified in
2. Historical context the failure report, but was noted as ‘‘in the vicinity of’’ the shell
to end-cap weld. Dawson noted that all the end caps were cold
In 1947 the concept of direct injection of liquid ammonia into formed and recommended using hot-formed or at least
soil as a nitrogen supplement was developed at Mississippi normalised heads. He concluded that ‘‘stress relieving of the
State College by Dr W.B. Andrews. This provided a 10-fold in- container as a whole should be considered the minimum quality
crease in yield of grain and cotton and in the late 1940s there pressure vessel for such service.’’ Loginow and Phelps (1961)
was a rapid growth in the use of anhydrous ammonia as an in- reported the findings of the AAI Research Committee indicat-
dustrial fertiliser (Prichard, 1998). Although ammonia had ing that air as a contaminant causes stress corrosion cracking
been in use for industrial applications for almost 100 years and that the tendency to stress corrosion is higher in higher
at that time, the use for transport and storage of agricultural yield strength steels, but that up to 0.1% water acts as an in-
ammonia of standard pressure vessels constructed in accor- hibitor. Loginow and Phelps made three recommendations
dance with ASME code requirements led to a significant num- for agricultural ammonia tanks: that tanks over 36 inches in
ber of failures in a short period of time. After a fatal accident diameter should be stress relieved during manufacture, that
involving the rupture of an ammonia holding tank a thorough care should be taken to eliminate air from tanks during oper-
study indicated that 3% of ammonia vessels were failing ation and that 0.2% distilled water by weight should be added
within three years of service. As this was a much higher inci- to the agricultural ammonia by the manufacturer. They con-
dence of failure than in other industries a research project was clude that ‘‘the water addition may be sufficiently effective that
started in 1954 by the Agricultural Ammonia Institute at stress-relieving may not be a necessary requirement at some future
Georgia Institute of Technology. This project identified stress date.’’ This seemed to draw the matter to a close and the ASM
corrosion cracking as the underlying cause of the high inci- Handbook reported that once the Department of Transport
dence of vessel failure in ammonia service, although it was had mandated the addition of 0.2% wt water to liquid ammo-
also noted that faulty workmanship had been a contributory nia ‘‘the problem of SCC of steel in ammonia has disappeared in the
factor in the fatal accident (Dawson, 1956). United States.’’ (Wilde, 1986).
Vessel failures due to stress corrosion cracking are also Despite this confidence there was continued research
found in the refrigeration industry, although the incidence is through the 1960s and 1970s into the causes of SCC in agricul-
much less than in the fertiliser industry (Bansch, 2007; Crack- tural ammonia systems, reflecting the remarks of Dawson in
nell, 1982; Pätz, 1995). About 80% of ammonia production out- 1956 that ‘‘there is a broad need for further research into the phe-
put is used in fertilisers and only 5% as a refrigerant (IIR, 1999). nomena of stress-corrosion cracking, and into the conditions of struc-
Also agricultural ammonia is a consumable process fluid, tural materials both metallurgically and chemically contributing to
whereas ammonia refrigerant is kept within a sealed system. their susceptibility to the phenomena known as stress-corrosion
Extensive reading and discussion with many industry experts cracking.’’ Much of the research, including the work of Logi-
in preparation for this paper suggests that there are signifi- now, Kim, Phelps, Wilde, Deegan and Jones, was conducted
cantly less than 10 leak incidents involving stress corrosion at the Applied Research Laboratories of the United States Steel
cracking in ammonia refrigeration each year in North America Corporation in Monroeville, PA. Phelps (1974) reported that ox-
and Europe. There are no known fatalities due to stress corro- ygen is not an inhibitor, and confirmed that stress corrosion
sion cracking in ammonia refrigeration vessels, and no cracking is not caused by hydrogen embrittlement, contrary
reported cases of vessels rupturing or fracturing in service. to the suggestion by Radd and Oertle (1971) that SCC was
Typically the indication of a leak due to stress corrosion crack- a form of hydrogen embrittlement. A comparative study of ag-
ing is a strong smell of ammonia gas seeping through a short ricultural ammonia containing 0.2% wt water with metallurgi-
hairline crack on a pressure vessel. The vessel can be isolated cal grade ammonia (Kim et al., 1975) confirmed that there was
and taken out of service without any risk of catastrophic fail- no sign of SCC in laboratory tests when water was present
ure, but of course the economic consequences for the end- whereas SCC occurred in every sample of air-contaminated
user, the installation contractor and the vessel manufacturer ammonia without water. The effect of electrochemical poten-
can be extremely damaging. tial was further investigated by Jones et al. (1977) confirming
that steels with a high free corrosion potential (in the range
400–600 mV against a pure aluminium reference) were more
3. Review of previous research susceptible to SCC and that cathodic protection or spraying
with zinc or aluminium offered protection against cracking.
The first report of the Agricultural Ammonia Institute’s Conversely, it has been established that anodic polarization,
Research Committee initial study was presented by Dawson giving a free corrosion potential above 400 mV versus an alu-
(1956), who remarked that ‘‘stress-corrosion cracking apparently minium reference, would produce stress corrosion cracking
is a little understood phenomena. It does not have a counterpart in samples that were oxygen-free, even if there was up to
nor is it similar to general corrosion which is identified as general 0.2% wt water in the ammonia (Nyborg and Lunde, 1996). A
wasting away of material; to pitting type corrosion which is identi- ‘‘unified mechanism of stress corrosion cracking’’ is described by
fied as pitting in localized areas; nor to general stress corrosion which Jones (1996). He identifies the formation of a passive surface
is identified as accelerated general corrosion and is attributable to layer as a prerequisite for the initiation of stress corrosion
744 international journal of refrigeration 31 (2008) 742–747

cracking. If the passive surface layer is ruptured then anodic the understanding of SCC in the agricultural industry and
dissolution of the metal below the rupture site causes rapid sought to apply the lessons learned to the refrigeration indus-
crack propagation in line with the original rupture. try. At that time, he said, ‘‘In spite of the long experience of using
A summary of the research conducted in the United States ammonia in the refrigeration industry there have been no published
of America in the 1970s is provided by Alexander and Laucks reports of leaks or cracks caused by SCC of the equipment and the
(2001). They review reports from the Southwest Research writer has only been shown one example and told of one other’’. In
Institute (SRI) and Battelle Pacific Northwest Laboratories, general the discussion following the paper agreed that there
both published in 1976, and conclude that the received wis- was no case history of SCC in refrigeration although a small
dom of that time was that high strength steel should be number of further examples were cited. Cracknell concludes
avoided, ammonia in refrigeration systems should contain that ‘‘it would be a wise precaution with new equipment to use only
at least 0.2% water and should be checked weekly, ‘‘extreme steels having specified minimum yield strengths below 350 N/mm2
care’’ should be taken to avoid air contamination and all com- and give a stress relief heat treatment after all welding and cold
ponents should be post-weld heat treated if possible. bending is complete.’’ The difficulty of providing access for inter-
Loginow, co-author of the original report from AAI, pro- nal inspection of refrigeration vessels, particularly shell and
vided a review of the 30 years of research that followed the tube condensers and evaporators, was highlighted.
original study (Loginow, 1986). He concluded that there were In recent years a number of cases of vessel leakage have
six practical means of avoiding SCC in agricultural storage been confirmed to be caused by SCC. This prompted the Inter-
tanks: purging air, dosing with water, avoiding reflux conden- national Institute of Ammonia Refrigeration (IIAR) to review
sation, applying post-weld heat treatment, using lower the common factors in these cases. The work of the IIAR
strength steel and inspecting the inner surfaces of the vessel SCC Task Force has provided a unique opportunity to review
frequently. Suggested inspection methods are wet fluorescent information from all parts of the world specifically in the con-
magnetic particle inspection (MPI) from the inside and acous- text of industrial refrigeration systems. Further information
tic emission testing from the outside. He identifies two bene- on the findings of the Task Force is provided by Bansch
fits of post-weld heat treatment; the reduction of residual (2007). This overview concludes that SCC is often the mecha-
stress and the reduction of surface hardness in the heat- nism of the start of the cracking process, but continued crack
affected zone around welds. Loginow is dismissive of pre- propagation is driven by more conventional failure mecha-
scribed limits for material yield strength, – ‘‘the formulation of nisms, such as high tensile stress, fatigue and stress concen-
a lower strength limit, below which SCC would not occur, is not prac- tration at crack tips. It also notes that ‘‘The surest approach to
tical because it depends on the level and nature of contamination, the eliminating SCC in ammonia refrigeration applications is to reduce
degree and nature of inhibition, the magnitude of residual and oper- the tensile stress by stress-relieving.’’
ating stresses and the operating temperature.’’ This view is sup-
ported by Nyborg and Lunde (1996) who state that ‘‘A
strength value below which SCC will not occur has not been identi- 4. Review of current codes and standards
fied, and in principle the lowest possible strength is the best. Some
companies have started to specify a maximum yield strength for The guidance available to the agricultural ammonia industry
the plates and not only an upper bound for the minimum yield as a result of the researches conducted in America, Europe
strength, in order to be able to discard the hardest plates in a batch’’ and Scandinavia is extensive. The International Association
It became apparent in the 1980s that stress corrosion af- of Classification Societies’ Guideline No. 33 (IACS, 1992) covers
fected many more ammonia storage facilities than had previ- ‘‘pressure vessel-type tanks’’ for transportation. It states that the
ously been thought (despite the optimism of the ASM nominal yield strength is to be less than 355 N/mm2 and the
Handbook), including low temperature (atmospheric) tanks. actual yield strength is to be less than 440 N/mm2. All tanks
Appl et al. (1989) reported on two atmospheric tank failures are to be post-weld heat treated, dished ends are preferably
in Europe, one at Seal Sands in England and one at Ludwigsha- to be hot-formed or alternatively normalised after cold form-
fen in Germany where over 300 stress corrosion cracks were ing and hardness of the weld, including the heat-affected zone
found in the tanks. Further research on low temperature tanks (HAZ), is specified to be not more than 230HV (Vickers hard-
was commissioned by a group of ammonia producers includ- ness scale). The European Fertiliser Manufacturer’s Associa-
ing BASF, DuPont, Norsk Hydro and ICI, and Nyborg and Lunde tion (EFMA) published a comprehensive guide in 2002 for
(1994) reported on laboratory simulation of low temperature atmospheric storage tanks (EFMA, 2002). It suggests that fa-
tank failures. They found that SCC was possible at 33  C tigue failure is not relevant to atmospheric tanks as there
but that cracks grew to roughly one-third of the depth of are not enough stress cycles in a typical 20 year tank life and
cracks at ambient temperature (þ18  C). They also tested sam- proposes a risk-based inspection regime for signs of SCC.
ples of material from the tanks at Seal Sands and Ludwigsha- Cracknell’s (1982) paper was used by the Institute of Refrig-
fen and reported (Nyborg et al., 1994) that the material from eration as the basis of an appendix to their Ammonia Safety
the cracked tanks ‘‘gave no indication of any unusual SCC suscep- Code in 1982 (Institute of Refrigeration, 1982) which states
tibility’’; it was not significantly different from the lab samples. ‘‘So far as is known SCC is less likely to occur on vessels that have
In the refrigeration community stress corrosion cracking been manufactured from low strength carbon steel (Re < 350 N/
was not reported prior to 1980, although it is probable that ves- mm2) and which have been furnace stress relieved’’. The IoR code
sel failures were occasionally occurring and being attributed was referenced extensively by the International Institute of
to other causes. Cracknell (1982) of I.C.I. presented a paper to Ammonia Refrigeration in preparing guidance for their Bulle-
the Institute of Refrigeration in London which summarised tin 110 in 1988 (IIAR, 1988) and the German Forschungsrat
 international journal of refrigeration 31 (2008) 742–747 745

Kältetechnik in 1995 (Pätz, 1995), but in all three cases the (A42CP) a higher grade of material was used (A48CP). SCC
guidance stopped short of specifying actual material proper- appeared in both vessels within a few months of service.
ties or mandating post-weld heat treatment. The German re- They had not been stress relieved as it was considered inap-
port concluded ‘‘No evidence was found in the literature on propriate for condensers. The actual yield stresses of five sam-
refrigeration technology and additionally in further journals that ples were in the range 352–379 N/mm2, although A48CP has
major failures of ammonia refrigeration systems were possibly con- a specified minimum yield strength of 285 N/mm2.
nected to SCC.therefore it can be concluded that no immediate re-
quirement for researching SCC exists, as faults in refrigeration 5.3. Case 3 – high pressure receiver, Holland
systems due to SCC have not been reported’’ (in translation). (no published information)
In the general pressure vessel standards, such as BS
PD5500 and ASME VIII, there is no specific reference to stress A high pressure receiver, manufactured in Germany in 2000,
corrosion in ammonia vessels, and no link between SCC and leaked shortly after being put into service. Cracks were found
yield strength, so no specific guidance on material selection between the end-cap knuckle and the weld, running perpen-
or specification of post-weld heat treatment for the avoidance dicular to the weld. The material strength is not known, but
of SCC is given to the designer. However, in the more recent the end caps were cold formed and the vessel had not been
European Standard EN14276:2006, paragraph 6.2 states ‘‘For post-weld heat treated.
vessels containing ammonia, stress corrosion cracking does not occur
if vessels are made from steel with ReH  360 N/mm2 and the refrig-
5.4. Case 4 – shell and tube evaporator, England
erating system is designed, operated and maintained according to
(no published information)
EN378-1–EN378-4’’ (CEN, 2006). However, there is no mention
of SCC in EN378:2000 parts 1–4, so the intention of the refer-
This case differs from the previous three: the affected vessel
ence in EN14276:2006 is not clear.
was on the low pressure side of the system, evaporating at
It is likely, in the light of recent re-examination of stress
2  C. Multiple SCC cracks were found in the vicinity of the
corrosion cracking data, for example in the paper by Bansch
shell to tubesheet weld, roughly at the normal liquid level.
(2007), that the code requirements contained in European,
The shell was fabricated from carbon steel with a yield
American and International standards for refrigeration safety
strength of about 370 N/mm2 and the vessel had not been
will be reviewed. A reduction in the reliance solely on good
post-weld heat treated.
purging and water content and a greater emphasis on post-
weld heat treatment are expected in the new safety standards.
5.5. Case 5 – suction header pipe on freezer, United
States of America (no published information)
5. Cases of stress corrosion cracking in
refrigeration This case is also different. The suction pipe of a spiral freezer
failed due to condensate-induced shock, but on inspection
Several brief cases of the occurrence of stress corrosion crack- evidence of weakening of the pipe due to stress corrosion
ing are presented here, and discussed in the following section. cracking was found. The cracks originated in tack welds on
In all cases the manufacturer and end-user are anonymous; a backing ring in the pipe. In this case the SCC was not the pri-
where a published report of the case is available it has been in- mary cause of failure, and it is likely that the shock would
dicated in the heading. Information on the cases was collected have caused a failure somewhere in the pipe: the SCC created
by the SCC Task Force of the International Institute of Ammo- a ‘‘weakest link’’ in the pipework.
nia Refrigeration in 2005 as part of their ongoing review of data.

5.1. Case 1 – high pressure receiver, England (National 6. Observations

Vulcan, 1988)
It is clear from the published literature and the case studies
A high pressure receiver, constructed in accordance with that the general ‘‘received wisdom’’ on SCC is not correct: it
BS5500, failed in service after three months of operation. An is not restricted to high pressure vessels, and it does not re-
ammonia leak was found from a transverse crack on the lon- quire significant amounts of oxygen in the ammonia; as little
gitudinal weld seam. The vessel was 914 mm diameter and as 0.5 ppm in laboratory tests can start the crack formation
5618 mm tan/tan, and was made from BS1501-151-430 plate (Lunde and Nyborg, 1991). While the addition of water is an ef-
10 mm thick. Extensive cracking was found along all of the fective inhibitor in storage ‘‘spheres’’ and agricultural nurse
longitudinal weld seams and around the knuckle in the end tanks, it provides no benefit in a refrigeration system. The
caps. The material yield strength was measured as 332 N/mm2 quantity of water in liquid ammonia will be about 650 times
and the vessel had not been post-weld heat treated. higher in equilibrium than the quantity in gaseous ammonia
above the liquid. This means that water will accumulate on
5.2. Case 2 – shell and tube condenser, France the low pressure side of the system, and will not offer protec-
(no published information) tion to the high pressure receiver. Conversely the quantity of
oxygen in gaseous ammonia is about 700 times higher in equi-
A pair of shell and tube condensers was installed in an ice- librium than the quantity dissolved in liquid, so oxygen will
cream factory. Although low yield strength steel was specified gather in the high pressure side. Yet, without any evidence
746 international journal of refrigeration 31 (2008) 742–747

of severe air contamination, SCC occurred in the evaporator in stress corrosion cracks are within the corrosion allowance for
case 4 and the suction header in case 5. the vessel and provided the conventional failure mechanisms
It is also clear that SCC occurs quickly when the vessel is are not causing further crack propagation the pressure vessel
put into service – probably within days if not hours of charging is likely to give many years of satisfactory service.
with ammonia – if the appropriate conditions prevail. It is not
generally detected until one of the cracks has grown in depth
to the full thickness of the shell, causing a leak. Crack growth 7. Conclusions
to full penetration can take much longer, typically between 6
and 24 months, but most of the cracks do not progress beyond There is no benefit from the addition of water to ammonia re-
about 1 mm in depth. Higher strength steel is significantly frigeration systems, as it will accumulate in the low pressure
more susceptible, and selection of steel with a low minimum side of the system. There is no chance of reaching a sufficiently
yield strength is not certain to prevent SCC as the actual low oxygen level through non-condensible gas purging to pre-
material strength may be much higher than the minimum. vent the formation of a passive monolayer on the inside of
High residual stress after welding makes SCC much more vessels. Therefore the only two ways of preventing the initia-
likely and no cases are reported of stress corrosion in vessels tion of stress corrosion cracks are removal of the stresses that
which were heat treated after fabrication. Highly stressed would cause the passive film to rupture and provision of a less
components, such as cold-formed end caps are also highly fragile (more ductile) surface layer. Both of these ends are
susceptible. High stresses in the tack welds of backing rings served by the post-weld heat treatment of vessels, which re-
seemed to be a contributory factor in case 5, and therefore it lieves residual stresses and provides a heavy oxide layer on
is concluded that backing rings should not be used for pipe- the inner surface of the vessel. The use of lower yield strength
work or vessels in ammonia refrigeration plants. It is not clear, steel reduces the chance of SCC occurring.
however, whether post-weld heat treatment is effective in The only feasible method of in-service inspection for SCC
preventing SCC because it reduces the residual stresses or be- in refrigeration vessels is ultrasound applied to the outer sur-
cause it modifies the inner surface of the vessel. There is no face of the vessel, with the sensitivity turned up far higher
evidence to suggest that thermal stresses are a factor in the than is usual for normal weld or material inspection. It should
onset of stress corrosion cracking, indeed the predominance be noted, however, that use of such high gain will detect im-
of high pressure receivers in the catalogue of known failures perfections in the weld metal and the shell that are within tol-
suggests that thermally induced stresses are not significant. erance and therefore not classed as defects. Ultrasound is
These observations suggest that stress corrosion cracking therefore a good method of tracking the development of
develops as the result of changes in the mechanical strength, SCC, but not so useful for initial identification. Where initial
ductility and corrosion potential of the inner surface of the microcracks (up to 1 mm deep) are detected they should be
vessel. The adsorption of oxygen on the inner surface may monitored regularly, but need not be repaired. Where a crack
produce a monolayer of material with lower ductility and a rel- has grown to (or close to) the point of leaking there is no
atively noble corrosion potential (Loginow, 1986). If the steel is appropriate repair method, and the vessel should be replaced.
relatively highly stressed, typically by the residual tensile
stress left after welding or cold forming, and if the steel has
a high yield strength then the ductility reduction may cause
8. Recommendations
fissures to appear in the monolayer normal to the direction
of tension. These are the passive film rupture sites described
Post-weld heat treatment should be applied wherever possible
by Jones (1996). The exposure of the more reactive steel sub-
to ammonia refrigeration pressure vessels, particularly those
strate material to the passive adsorbed monolayer will then
that operate at temperatures above 5  C or those that may
cause galvanic corrosion to progress into the substrate in
be out of service for prolonged periods. It is also imperative
line with the original fissure, creating a mix of transgranular
that the actual yield strength of the material should be as low
and intergranular cracks with the feathered ends typical of
as possible; preferably below 325 N/mm2 and definitely below
SCC. The development of these cracks will be arrested if the
350 N/mm2. Unfortunately the suggestion of Nyborg and Lunde
exposed surface is re-filmed. Loginow (1986) suggests that
(1996) of specifying a maximum yield strength does not appear
this is the inhibiting mechanism of water in the ammonia,
to be a practical proposition for the refrigeration industry.
and can also be achieved if excess oxygen is present, but the
Backing rings should not be used for pipe or vessels in ammo-
re-filming will be less effective if nitrogen is also present.
nia service under any circumstances. Further study of the com-
It is concluded that high strength steel is more susceptible
position, yield strength and ductility of the passive film
to SCC because it is more ductile than the adsorbed oxygen
described here is warranted. It is also possible that additional
passive layer. Lower strength steel tends to yield with the pas-
research will identify an alternative surface treatment to be
sive layer, reducing the probability of rupture sites forming as
applied in cases where post-weld heat treatment is not feasible.
the surface ductility reduces.
Bansch (2007) suggests that crack propagation in industrial
systems is due to more conventional failure mechanisms. This
leads to the conclusion that SCC ‘‘micro-cracks’’ up to 1 mm Acknowledgments
deep will be found in vessels where a crack has propagated to
failure, and furthermore that ‘‘micro-cracks’’ are also likely The author would like to thank his co-workers at Star Refrig-
to be present in vessels which have not failed. Provided the eration, the International Institute of Ammonia Refrigeration
 international journal of refrigeration 31 (2008) 742–747 747

and the University of Strathclyde for their support and contri- International Institute of Refrigeration, 1999. Ammonia as
butions to this study. Thanks are also due to the directors of a Refrigerant. IIR, Paris, p. 21.
Star Refrigeration for funding the study and for permission Jones, D.A., 1996. Principles and Prevention of Corrosion, second
ed. Prentice Hall, Upper Saddle River, NJ.
to publish the results.
Jones, D.A., Kim, C.D., Wilde, B.E., 1977. The electrochemistry and
mechanism of stress corrosion cracking of constructional
steels in liquid ammonia. Corrosion 33 (2), 50–55. National
references Association of Corrosion Engineers, Houston, TX.
Kim, C.D., Wilde, B.E., Phelps, E.H., 1975. Stress corrosion cracking
of line-pipe steels in anhydrous ammonia. Corrosion 31 (7).
Alexander, V., Laucks, M., 2001. Stress corrosion cracking of steels National Association of Corrosion Engineers, Houston, TX.
in ammonia. In: Proceedings of the International Institute of Loginow, A.W., 1986. Stress-corrosion cracking of steel in
Ammonia Refrigeration, Long Beach, CA, pp. 365–378. liquefied ammonia service – a recapitulation. Materials
Appl, M., Fässler, K., Fromm, D., Gebhardt, H., Portl, H., 1989. New Performance 25 (12). National Association of Corrosion
cases of stress corrosion cracking in large atmospheric Engineers, Houston, TX.
ammonia storage tanks. In: Proceedings of AIChE Safety in Loginow, A.W., Phelps, E.H., 1961. Stress-corrosion Cracking of
Ammonia Plants and Related Facilities Symposium, San Steels in Agricultural Ammonia. American Institute of
Francisco, CA. Chemical Engineers, Lake Placid, NY.
Bansch, R., 2007. Stress corrosion cracking in the ammonia Lunde, L., Nyborg, R., 1991. Stress Corrosion Cracking of Carbon
refrigeration industry. In: Proceedings of the International Steel Storage Tanks for Anhydrous Ammonia. The Fertiliser
Institute of Ammonia Refrigeration, Nashville, TN. Society, London.
CEN, 2006. EN14276 – 1 Pressure equipment for refrigerating Nyborg, R., Lunde, L., 1994. Stress corrosion cracking in low
systems and heat pumps – Part 1: Vessels – general temperature ammonia storage tanks. In: Proceedings of AIChE
requirements, Brussels, 2006. Ammonia Safety Symposium, Vancouver.
Cottis, R.A., 2000. Stress Corrosion Cracking: Guide to Good Nyborg, R., Lunde, L., 1996. Measures for reducing SCC in
Practice in Corrosion Control. National Physical Laboratory, anhydrous ammonia storage tanks. Process Safety Progress 15
Teddington, UK. (1). American Institute of Chemical Engineers, NY.
Cracknell, A.J., 1982. Stress corrosion cracking of steels in Nyborg, R., Drønen, P., Lunde, L., 1994. SCC in low-temperature
ammonia. In: Proceedings of the Institute of Refrigeration, ammonia storage tanks. In: Proceedings of AIChE Ammonia
London, vol. 78, pp. 70–75. Safety Symposium, Vancouver.
Dawson, T.J., 1956. Behavior of welded pressure vessels in Pätz, G., 1995. Sicherheitsrelevante Werkstoffprobleme –
agricultural ammonia service. The Welding Journal, New Literaturstudie zur Bewertung der Spannungsrißkorrosion
York, NY, 568–574. von Werkstoffen in Ammoniak-Kälteanlagen. Forschungsrat
EFMA, 2002. Recommendations for the Safe and Reliable Inspection Kältetechnik, Frankfurt.
of Atmospheric, Refrigerated Ammonia Storage Tanks. Phelps, E., 1974. Causes of stress corrosion cracking of steel in
European Fertilizer Manufacturers’ Association, Belgium. ammonia, Ammonia Plant Safety, Vol. 16.
IACS, 1992. Guidelines for the Construction of Pressure Vessel Prichard, J.G., 1998. Making Things Grow: The Story of Mississippi
Type Tanks Intended for the Transportation of Anhydrous Chemical Corporation. University Press of Mississippi,
Ammonia at Ambient Temperatures. International Jackson, MS.
Association of Classification Societies. Guideline No. 33. Radd, F.J., Oertle, D.H., 1971.Getter-ion probe studies of hydrogen
Institute of Refrigeration, 1982. A Safety Code for Compression entry into iron from liquid and gaseous ammonia systems. In:
Refrigerating Systems Utilizing Ammonia Part II – Proceedings of National Association of Corrosion Engineers,
Commissioning, Inspection and Maintenance. Institute of Chicago, IL.
Refrigeration, London. Wilde, B., 1986. Stress-corrosion cracking, second ed. In: ASM
International Institute of Ammonia Refrigeration, Bulletin 110, Handbook, vol. 11: Failure Analysis and Prevention. ASM
Annex E, Chicago, IL, 1988. International, Materials Park, OH, p. 215.