Engineering Failure Analysis 18 (2011) 1403–1406

Contents lists available at ScienceDirect

Engineering Failure Analysis

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

Short communication

An overview of naphthenic acid corrosion in a vacuum distillation plant

Paulo P. Alvisi a,1, Vanessa F.C. Lins b,⇑
a
PETROBRAS Industry, Fernão Dias, BR 381, km 427, 32530-000, Betim, Brazil
b
Federal University of Minas Gerais, Antonio Carlos Avenue 6627, Belo Horizonte, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: After 5 years processing about 6500 m3/day of reduced crude oil with an average total acid
Received 17 November 2010 number of 1.2 mg KOH/g and total sulfur content of 0.8 wt%, a vacuum distillation plant
Received in revised form 19 March 2011 had its operations interrupted to suffer maintenance and inspection. Following theoretical
Accepted 28 March 2011
considerations, this work shows the inspection outcome and tries to stress the several mor-
Available online 1 April 2011
phologies resulting from the naphthenic acid attack as well as different types of action by
which the naphthenic acid corrosion occurs in a vacuum distillation plant.
Keywords:
Ó 2011 Elsevier Ltd. All rights reserved.
Chemical-plant failures
Naphthenic acid corrosion
Petrochemical
Vacuum distillation plant

1. Introduction

The most vulnerable unit of naphthenic acid corrosion (NAC) in a crude oil refinery is the vacuum distillation plant.
Although naphthenic acids with several molar masses may be present in crude oil, in general they tend to concentrate in
the range with molar mass that corresponds to hydrocarbon fractions with boiling points coincidental to those which are
found in vacuum plants [1]. This fact occurs in heavy vacuum gas oil (HVGO), which is the fraction that normally acquires
the highest total acid number (TAN).
The reliability of such plants is strongly dependent on their metallurgy. Their critical points as furnace tubes and elbows,
transfer line, walls, and trays of vacuum column must be built with Mo-bearing stainless steel such as AISI 316 and AISI 317
stainless steels [2]. Under certain circumstances and at given molybdenum content (particularly for the AISI 316 steel with
less than 2.5 wt%), none of these materials can resist the NAC attack, therefore molybdenum richer alloys are necessary.

2. Theory: mechanism and interfering variables

The chemical reaction that is known to represent NAC to iron alloys is:

Fe þ 2R½CH2 n COOH ¼ FeðR½CH2 n COOÞ2 þ H2 ð1Þ

where R denotes the naphthenic radical (one or more cyclopentane or cyclohexane saturated ring) linked to carboxyl groups
through n (typically >12) units CH2, and the corrosion products, the complex Fe(R[CH2]nCOO)2, is soluble in oil. It must be
observed that the acid tends to dissociate as R[CH2]nCOO– and H+ even in an organic liquid environment [2,3].

⇑ Corresponding author. Tel.: +55 31 34091775; fax: +55 31 34091726.

E-mail addresses: paulopa@petrobras.com.br (P.P. Alvisi), vlins@deq.ufmg.br (V.F.C. Lins).
1
Tel.: +55 31 35294584; fax: +55 31 34091789.

1350-6307/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.engfailanal.2011.03.019
1404 P.P. Alvisi, V.F.C. Lins / Engineering Failure Analysis 18 (2011) 1403–1406

Table 1
Corrosion rate in function of temperature and total acid number.

Corrosion rate Total acid number Temperature (°C)

(mm/y) (mg KOH/g)
0.30–0.58 2–10 204
0.45–0.86 2–10 232
0.59–1.08 2–10 260
0.74–1.38 2–10 288
0.89–1.63 2–10 316
1.00–1.85 2–10 345
1.27–2.08 2–10 371

Considering a surface free of protective films, the reaction above implies that a direct contact of the ligand, R[CH2]nCOO–
with an alloy system such as FeCrNiMoMn (AISI 316 and 317 stainless steels) causes iron corrosion. Such fact was shown by
the determination of crystal field stabilization energy (CFSE) that is derivative from ligand field theory which explains the
nature of chemical bonding in transition-metal complexes.
It was verified that Fe and Fe2+ when compared with the atoms Cr, Ni, Mo, Mn, and its ions, Cr3+, Ni2+, Mo3+, Mn2+, have the
strongest affinity with the ligand R[CH2]nCOO– considering that the more negative is the crystal field stabilization energy
(CFSE), and the more stable is the complex [2]. It was concluded that the Fe(R[CH2]nCOO)2 corresponds to more stable com-
plexes which are formed from ‘‘Fe-rich active spots’’ offering an explanation about the vulnerability of Fe base alloys in rela-
tion to the naphthenic acids [2]. Spots could be present on the surface film in points where there are faults.
The molybdenum produces a benefic influence on the general characteristics of the surface film due to an advantageous
synergism that exists between chromium and molybdenum. Molybdenum increased the stability of the passive layer [4,5].
Thus, all these facts could justify the role of Mo in the behavior of austenitic stainless steel in crude oil environments [2].
Wu et al. [2] reported a decrease from 1.8  103mm/year of corrosion rate to 0.6  103mm/year for an austenitic stainless
steel without Mo and a steel with a content of 7 wt% of molybdenum. The variables in the work of Wu et al. were a total acid
number of 210 mg KOH/g, temperature of 250 °C, and 500 h of testing.
Both total acid number and temperature exert an influence on naphthenic acid corrosion. Although the intrinsic aggres-
siveness of a specific type of naphthenic acid can vary with the molecular weight and with molecule complexity, generally
the increase of total acid number (TAN) is associated with increasing corrosivity. In relation to the temperature, the field expe-
rience showed that NAC becomes important only above 200 °C and ends its aggressiveness at about 400 °C, possibly because
of the thermal decomposition of the acid. It is believed that temperature increases diffusion transport of acid toward the metal
surface, the dissolution of corrosion products and aggravates the corrosion reaction due to its endothermic nature [2,3].
Considering NAC on carbon steel, for a same TAN, the corrosion rate increases with increase of temperature, and at the
same temperature, the corrosion rate increases with TAN. Piehl [1] and Slavcheva et al. [3] show that as temperature in-
creased from 204 °C to 371 °C, the corrosion rate increased from 0.30 mm/y to 1.27 mm/y (2 mg KOH/g). At 371 °C, the cor-
rosion rate increased from 1.27 mm/y to 2.08 mm/y as the total acid number increased from 2 mg KOH/g to 10 mg KOH/g as
shown in Table 1.
In a refinery, a given hydrocarbon fraction always contains in some level sulfur compounds with different chemical affin-
ities to Fe. For example, the sulfur sulfide (H2S) can produce superficial films (FeS):

Fe þ H2 S ! FeS þ H2 ð2Þ
The FeS film protects the steel from NAC. The same mechanism occurs with mercaptans (R–SH). However, the film will
not be formed by the interaction with compounds like thiophene (C4H4S) [3,6]. A dangerous situation may occur if the sulfur
compound were able to produce water as it occurs in Eq. (3):

R2 S ¼ O þ H2 ! R2 S þ H2 O ð3Þ
The hydrogen is produced by the corrosion reaction and the sulfur compounds produced are sulfoxides (R2S = O). In this
case the water could provide an additional (besides the own liquid hydrocarbon) media where naphthenic acid can disso-
ciate [7].
The naphthenic acid corrosion is also strongly affected by flow regime. Being a liquid phenomenon, in a two phase flow
with a given degree of vaporization (DV), such as occurs in furnace tubes, outlets headers, and transfer line, the NAEC sever-
ity will depend on unvaporized residue TAN. In addition, recognizing the importance of the wall shear stress in removal of
the sulfide protective film (if it exists), it was verified that such stress reduces with a degree of vaporization but increases
along fluid velocity (for a given DV). Velocity and shear stress seem to exert no effect on NAEC when the formation of drop-
lets of naphthenic acids condensate in a tower environment [1,6].

3. Material and methods

The inspection method consists of visual observation and analysis by using optical and scanning electron microscopy. The
alloys were analyzed by using the X-ray fluorescence spectrometer Niton XLi 818, source Am241.
 P.P. Alvisi, V.F.C. Lins / Engineering Failure Analysis 18 (2011) 1403–1406 1405

Fig. 1. Different kinds of corrosion morphologies connected to the naphthenic acid attack. Region 1 is the IMPT random packing, region 2 is the tray and
bubble caps, region 3 is the column wall flash zone, region 4 is the support grid, and region 5 is the transfer line.

The morphology of material was determined using the scanning electron microscopy, EVO 50 Zeiss equipment, 20 kV of
acceleration voltage, coupled with the energy dispersive spectrometer, INCA 350 model.
1406 P.P. Alvisi, V.F.C. Lins / Engineering Failure Analysis 18 (2011) 1403–1406

4. Results and discussion

The inspection was accomplished giving priority to certain positions where the damage caused by NAC was considered
more likely (literature, field experience and component metallurgy).
The transfer line and internal walls of the vacuum column suffered damages. The tower was attacked mainly in the com-
ponents where heavy vacuum gas oil (HVGO), with the total acid number unknown, was prevailing, as opposed to what oc-
curs to positions correspondent to light vacuum gas oil (LVGO). However, nothing was found in the HVGO–LVGO side cut
piping with pipe in AISI 316 stainless steel and carbon steel respectively.
The temperature range between the furnace and the flash zone is from 365 °C to 394 °C. The average temperature in the
flash zone is 365 °C, in the HVGO is 347 °C, and in LVGO is 282 °C.
The total acid number in the region of heavy vacuum gas oil (HVGO) was higher than the value of reduced crude oil
(1.2mgKOH/g). The values of total acid number (TAN) of liquid residue in the transfer line and of TAN of hydrocarbons in
the upper region of tower are unknown.
The inspection was performed after five years of operation.
Region 1 in Fig. 1 is the Intalox Metal Tower Packing (IMPT), a random packing whose objective is to impose pipe resis-
tance (charge loss) to the gas flow in order to liquefy and separate the heavy fractions. Pitting attack promoted by acid con-
densate occurred in materials with thickness of 0.3 mm. In particular, thin components should be specified with higher Mo-
bearing stainless steel as the AISI 316 stainless steel with >2 wt% Mo.
Region 2 corresponds to the tray and bubble caps. The naphthenic acid corrosion process was clearly assisted by mechan-
ical component, and friction developed between bubble caps free to move relatively to the tray. NAC process was developed
at 350 °C. The bubble caps caused a corrosive/fretting attack. The motion between the bubble caps and tray removes protec-
tive films and results in accelerated attack. Also, the iron oxides/hydroxides that are produced as corrosion products are
abrasive and increases the removal of protective films and in direct abrasion of the metal. The tray was perforated after five
years, and the initial thickness was 3 mm. The material of tray and bubble caps is the AISI 316 stainless steel with >2 wt% Mo.
Region 3 is the column wall flash zone. The average temperature in the flash zone is 365 °C. Naphthenic acid corrosion
occurred, assisted by impingement occasioned by an expansion fluid. The adjacent surface of steel is protected by a black
sulfide film. The material of the column wall is the AISI 410 stainless steel (3 mm in thickness).
Region 4 is the support grid with thickness of 5 mm. This is a typical case where the NAC pitting attack was promoted by
acid condensation and the fluid velocity did not exert any effect. Naphthenic acid corrosion occurred at 355 °C. The material
is the AISI 304 stainless steel.
Region 5 is the transfer line. This is a damage produced by NAC (at 380 °C) where the flow regime strongly affects the fluid
corrosivity. The metallurgy of material is 5Cr–0.5Mo (wt%) alloy steel. The thickness of pipes is 12 mm.

5. Conclusion

Naphthenic acid corrosion is a highly complex phenomenon involving different variables. As shown, such complexity is
apparently reflected in the different morphologies produced by naphthenic attack on different steel and conditions. Pitting
attack promoted by acid condensate, naphthenic acid corrosion assisted by mechanical component (friction and impinge-
ment), and naphthenic acid corrosion where the flow regime affects the fluid corrosivity occurred in a vacuum distillation
plant after it had been processed, for five years, reduced crude oil with a total acid number of 1.2 mg KOH/g and a total sulfur
of 0.8 wt%.

References

[1] Piehl RL. Naphthenic acid corrosion in crude distillation units. Mater Perform 1988;27:37–43.
[2] Wu XQ, Jing HM, Zheng YG, Yao ZM, Ke W. Resistance of Mo-bearing stainless steels and Mo-bearing stainless-steel coating to naphthenic acid corrosion
and erosion–corrosion. Corr Sci 2004;46:1013–32.
[3] Slavcheva E, Shone B, Turnbull A. Review of naphthenic acid corrosion in oil. refining. Br Corr J 1999;34:125–31.
[4] Sekine I, Kawase T, Kobayashi M, Yuasa M. Corros Sci 1991;32:815–25.
[5] Ahn MK, Kwon HS, Lee HM. Corros Sci 1998;40:307–22.
[6] Tebbal S, Kane RD. Review of critical factors affecting crude corrosivity, Corrosion’96, 1996, paper no. 607.
[7] Yépez O. Influence of different sulfur compounds on corrosion due to naphthenic acid. Fuel 2005;84:97–104.