Journal of Environmental Management 151 (2015) 105e112

Contents lists available at ScienceDirect

Journal of Environmental Management

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

A comparative study of the treatment of ethylene plant spent caustic

by neutralization and classical and advanced oxidation
Alaa Hawari a, *, Hasanat Ramadan b, Ibrahim Abu-Reesh c, Mabrouk Ouederni d
a
Department of Civil and Architectural Engineering, College of Engineering, Qatar University, P.O. Box 2713, Qatar
b
Environmental Engineering Master Program, College of Engineering, Qatar University, P.O. Box 2713, Qatar
c
Department of Chemical Engineering, College of Engineering, Qatar University, P.O. Box 2713, Qatar
d
Qatar Petrochemical Company (QAPCO), P.O. Box 756, Qatar

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

Article history: The treatment of spent caustic produced from an ethylene plant was investigated. In the case of
Received 23 October 2014 neutralization alone it was found that the maximum removal of sulfide was at pH values below 5.5. The
Received in revised form higher percentage removal of sulfides (99% at pH ¼ 1.5) was accompanied with the highest COD removal
20 December 2014
(88%). For classical oxidation using H2O2 the maximum COD removal percentage reached 89% at pH ¼ 2.5
Accepted 22 December 2014
Available online 26 December 2014
and at a hydrogen peroxide concentration of 19 mM/L. For the advanced oxidation using Fenton's process
it was found that the maximum COD removal of 96.5% was achieved at a hydrogen peroxide/ferrous
sulfate ratio of (7:1).
Keywords:
Spent caustic
© 2014 Elsevier Ltd. All rights reserved.
Neutralization
Oxidation
Advanced oxidation
Ethylene plant

1. Introduction
Sodium hydroxide solutions are used in many chemical in-
dustries (i.e. liquefied petroleum gas, natural gas, and refining in-
dustries to wash out different gases such as hydrogen sulfide and
carbon dioxide from different hydrocarbon streams (Hashemi and
Heidarinasab, 2012). Once these gases react with sodium hydrox-
ide a waste solution known as spent caustic will be produced
(Hashemi and Heidarinasab, 2012). Spent caustic is considered to
be one of the liquid industrial wastes that are not easy to handle
and to dispose of, due to its high content of pollutants, alkalinity
(pH > 12), high salinity (sodium of 5e12 wt%) and high sulfide
concentration (2e3 wt%) (Kumfer et al., 2010; Alnaizy, 2008;
Olmos, 2004; Sheu and Weng, 2001). Accordingly, special man-
agement of spent caustic is required where treatment of spent
caustic before it could be sent to a conventional wastewater
treatment plant is a must.
Depending on which industry is producing the spent caustic and
on the source of fuel that the fresh caustic is washing, spent caustic
can be classified into three main types. Table 1 summarizes the
* Corresponding author.
E-mail address: a.hawari@qu.edu.qa (A. Hawari).
three main types of spent caustic and their main characteristics.
Usually refineries do not separate each type of spent caustic and
they mix the three types, this is referred to as the mixed refinery
spent caustic (Alnaizy, 2008). Numerous efforts have been made to
develop and to enhance the treatment process of spent caustic.
Treatment methods for spent caustic can be classified into three
main categories: biological, chemical and thermal processes
(Ahmad, 2010). Every category has its own advantages and disad-
vantages. Among the three mentioned categories chemical treat-
ment processes are considered to be the most common. Chemical
oxidation can be classified into two categories: classical chemical
oxidation and advanced chemical oxidation processes (AOPs). In
the classical chemical oxidation process a chemical oxidant is
directly added to the waste stream in order to oxidize any available
contaminants. The most commonly used chemical oxidants are:
chlorine, chlorine dioxide, oxygen, persulfate, permanganate,
ozone, and hydrogen peroxide. Table 2 summarizes the main ad-
vantages and disadvantages of each oxidant.
Advanced chemical oxidation processes (AOPs) are considered
to be promising methods for the treatment of spent caustic. AOPs
would form sufficient quantities of highly reactive hydroxyl radicals
(HO) at near ambient temperature and pressure in order to attack
complex chemical contaminants in the waste stream (Munter,

http://dx.doi.org/10.1016/j.jenvman.2014.12.038
0301-4797/© 2014 Elsevier Ltd. All rights reserved.
106 A. Hawari et al. / Journal of Environmental Management 151 (2015) 105e112

Table 1 2. Materials and methods

Different spent caustic types and their characteristics.

Type of spent Sulfidic Cresylic Naphthenic Ref. 2.1. Characterization of spent caustic
caustic

Source Ethylene and Gasoline Kerosene and Diesel (Kumfer
Liquefied
Petroleum Gas
(LPG)
Content High High High concentration of (Kumfer
concentration concentration polycyclic aliphatic et al.,
of sulfides and of phenols & organic compounds 2010)
mercaptans cresols
Chemical 5000e90,000 50,000 150,000e240,000
Oxygen e100,000
Demand
(COD)
(ppm)
Total Organic 20e3000 10,000 24,000e60,000
Carbon e24,000
(TOC)
(ppm)
Sulfides 2000e52,000 <1 0e63,000
(ppm)
Total phenol 2e30 1900e1000 14,000e19,000
(ppm)
2001). AOPs could reduce the concentration of certain pollutants in
spent caustic to a very low concentration below 5 ppb (Canizares
et al., 2009). In order to achieve the required level of treatment
AOPs are usually paired with other treatment systems. In such cases
the treatment process will be more complex and more costly. A
large number of methods are classified as AOPs. The generation of
the hydroxyl radicals are achieved by the use of one or more strong
oxidants (e.g. H2O2, O2, and O3) and/or catalysts (e.g. titanium di-
oxide, transition metal ions) and/or energy sources (e.g. ultraviolet
radiation) (Munter, 2001). The selection of a certain advanced
oxidation process depends on the type of compounds to be
removed, treatment objectives, concentrations of pollutants, site
considerations, and cost. The main processes found in literature for
producing the required radicals are summarized in Table 3.
It can be noted from Table 3 that the UV system has major
drawbacks such as mass transfer limitations and turbidity and
other compounds (i.e. nitrate) interference with UV light which will
result in lowering the process efficiency. The ozone/hydrogen
peroxide system is similar to UV/hydrogen peroxide system in
many ways. However, the ozone/hydrogen peroxide system is less
affected by the feed characteristics. The combination of ozone,
hydrogen peroxide, and ultraviolet radiation (O3/H2O2/UV) in one
system will enhance the treatment process. However, in this case
the cost of treatment will be high.
As mentioned before spent caustic solutions generated from
petroleum refineries and/or from petrochemical industries vary
highly in quantity and quality. These depend on the type of pro-
cessed material, the different operational units, and the different
operational conditions. In this study spent caustic produced from
an ethylene plant was targeted. Ethylene is produced from an
ethane rich gas (ERG) which is supplied from Qatar petroleum re-
finery. None of the previous studies compared the treatment of
spent caustic produced from an ethylene plant by several chemical
processes namely, neutralization, neutralization coupled with
classical oxidation, neutralization coupled with advance oxidation.
The targeted COD and sulfide concentrations were 1000 mg/L and
2 mg/L, respectively. At these concentrations the treated spent
caustic can be handled by conventional biological treatment. The
effect of different parameters on the treatment process was
investigated namely, pH, oxidants concentration and the ratio of
oxidant concentration to catalyst concentration.
Spent caustic is a dark brown to black solution its composition is
et al., highly variable depending on the type of petrochemical industry it
2010) is produced from. Samples of spent caustic were collected from
Qatar Petrochemical Company (QAPCO). QAPCO is a petrochemical
company in Qatar that produces ethylene and polyethylene for the
plastic industry. Three samples were collected from QAPCO's
petrochemical plant at three different times. After collection sam-
(Ahmad,
ples were stored in a refrigerator at 4  C until use. Each test was
2010)
done in triplicate. The main characteristics of the collected samples
are summarized in Table 4. In Table 4 the range (min. value e max.
value) of the analyzed samples is presented.
(Ahmad, Comparing the results in Table 4 with other refinery spent
2010)
caustic solution it was found that ethylene spent caustic has a
relatively low COD and BOD concentrations. Where the COD and
(Ahmad, BOD concentrations for other refinery spent caustic solutions range
2010) between (20,000e50,000 ppm) and (5000e15,000 ppm), respec-
(Ahmad, tively. Moreover, sulfide and phenol concentrations are also rela-
2010)
tively lower than other refinery spent caustic. Where the sulfide
and phenol concentrations for other refinery spent caustic solu-
tions range between (0e64,000 ppm) and (0e20,000 ppm),
respectively. The different characteristics of such spent caustic is
expected to highly affect the treatment process.
2.2. Experimental procedure
Neutralization and Oxidation experiments were performed in a
batch system with different reagents concentrations. The experi-
mental apparatus used in the batch tests consisted of a magneti-
cally stirred round bottom pyrex glass flask (250 ml), 25 ml burettes
were used to add the required reagents, and temperature was
controlled by a magnetic stirrer hot plate (Stuart Scientific, UK).
Temperature and pH were continuously measured using a pH/ORP
meter (HANNA Instruments, HI 2211). In order to ensure homoge-
nous conditions in the reactor the solution was continuously mixed
at a speed of 100 rpm. The degree of stirring was kept mild as any
excessive stirring lead to excessive foaming. Sulfuric acid 98% (v/v)
(panreac) and 5.0 M sodium hydroxide solutions were used for pH
control. Lab grade hydrogen peroxide at the concentration of 30%
(v/v) (panreac) was used as the oxidant reagent and ferrous sulfate
(heptahydrated) was used as the catalyst. The advanced oxidation
process was implemented by adding ferrous sulfate catalyst
(6.6 mM/L), then hydrogen peroxide (20 mM/L) was dosed into the
reactor. After 60 min of reaction time samples were withdrawn
from the reactor, neutralized to a pH value around 7.0, centrifuged
to separate the iron floc at a speed of 2000 rpm using a Centurion
Scientific centrifuge (K3 series) and then samples were decanted
and analyzed for COD and sulfide concentrations.
3. Results and discussion
3.1. Neutralization
Fig. 1 shows the impact of pH value on the sulfide and COD
removal. From Fig. 1 it could be seen that as pH value decreases the
percentage removal for both sulfides and COD increases. The reason
behind the increase of sulfides removal at low pH values is that
sulfides exist in three different forms H2S, HS, S2 depending on
the pH value as shown in reactions (1) and (2):
 A. Hawari et al. / Journal of Environmental Management 151 (2015) 105e112
H2 S4HS þ Hþ pKa ¼ 7:1
HS 4S2 þ Hþ pKa ¼ 14
At a pH value equal to 7.0 approximately 50% of the total sulfides
is in the form of H2S gas while the other 50% is present as HS. As
pH value decreases it would be expected that most of the total
sulfides will be in the form of H2S gas which will be released into
the atmosphere. Also, as pH value increases it would be expected
that most of the total sulfides will be in the form of HS dissolved in
solution (Miessler, 1991). Therefore, it would be expected that
removal of sulfides in the form of H2S gas would increase as pH
value decreases. As shown in Fig. 1 maximum removal of sulfide
was at pH values below 5.5. Decreasing the pH value below 5.5 had
a minimal effect on the removal percentage of sulfide where at pH
values of 1.5, 3.5 and 5.5 almost similar removal percentages were
obtained. Similar results were obtained by several researchers
where it was found that neutralization (pH value between 4 and 5)
of spent caustic reduced sulfide concentration significantly (Sheu
and Weng, 2001). It was noticed in this study that as the pH
value starts decreasing excessive foaming was formed and high
odors were produced which is an indication of the forming of H2S
gas. Also, the foaming could be due to the presence of polycyclic
aliphatic organic compounds and cresols (Kumfer et al., 2010). In
addition to foaming an extreme color change was observed an
indication of the chemical reactions occurring in the system.
From Fig. 1 it was also observed that the decrease of sulfides was
accompanied by a decrease of COD. The higher percentage removal
of sulfides (99% at pH ¼ 1.5) was accompanied with the highest COD
removal (88%). This could be due to the removal of sulfide COD. The
remaining COD is due to the remaining non-sulfide compounds
which could be mainly due to emulsified hydrocarbons (Sheu and
Weng, 2001). Although the removal percentages reached very
high values for both sulfides and COD still the remaining concen-
trations in solution were still higher than the desired limits of
discharge. The remaining concentrations for sulfides and COD were
9.9 mg/L and 1800 mg/L, respectively. According to the industrial
case study the required limits in order to discharge the treated
spent caustic effluent into a secondary bioreactor were 2 mg/L
sulfides and 1000 mg/L COD. Therefore, it was decided that
neutralization alone is not enough. Neutralization coupled with
classical oxidation was performed.
3.2. Neutralization coupled with classical oxidation
The reaction of hydrogen peroxide in a spent caustic solution is
considered to be a complex reaction which will be affected by
different variables, such as pH value, initial concentration of
hydrogen peroxide, and the time of the reaction. These variables
control the rate of the reaction, the consumption of hydrogen
peroxide and the end products formed (Ahmad, 2010). Hydrogen
peroxide is a multipurpose oxidant for many systems. It can be
applied to the system directly or with a catalyst. Incase catalyst is
employed then this process is known as an advanced oxidation
process (AOP). The effect of hydrogen peroxide (H2O2) concentra-
tion on COD removal in the spent caustic solution was studied.
Different hydrogen peroxide concentrations were added to the
spent caustic samples at different pH values and for a 60 min re-
action time. As shown in Fig. 2, at the beginning when the spent
caustic pH value was adjusted to a value of 1.5, 2.5, and 3.5 the COD
removal reached a value around 87%, 84%, and 80%, respectively. As
H2O2 was added to the system, it was noticed that the COD removal
percentage starts to increase up to a certain point where after that
point the removal percentage starts decreasing. The maximum COD
107
removal percentage reached 89% at pH ¼ 2.5 and at a hydrogen
(1) peroxide concentration of 19 mM/L. Above this concentration COD
removal started to decrease as the concentration of hydrogen
(2) peroxide increased. The COD removal increases as the concentra-
tion of hydrogen peroxide increases due to the generation of hy-
droxyl radicals. Hydroxyl radicals tend to further oxidize organic
pollutants. At high concentrations of hydrogen peroxide, hydrogen
peroxide tends to react with the hydroxyl radicals and produce a
weaker radical HO2 as shown in reaction (3) (Kang et al., 1999). As a
result, hydrogen peroxide will not be fully utilized to generate
hydroxyl radicals but it will convert some of the active hydroxyl
radicals into weaker radicals.
OH$ þ H2 O2 /H2 O þ HO2 $ (3)
As shown in Fig. 2 even at a very low concentration of H2O2
(2 mM/L) sulfides were totally removed from solution. H2O2 can
control sulfides by destruction by oxidizing sulfides to elemental
sulfur or sulfate ions. The solution's pH value will highly affect the
reaction between sulfides and hydrogen peroxide. At low pH
values, sulfide exists mainly as hydrogen sulfide, which will react
with hydrogen peroxide on a 1:1 (w/w) basis to form elemental
sulfur. Therefore, the best usage of hydrogen peroxide would be at
low pH values. For this reason the experiments were performed at
such low pH values. At a pH value around 7 (neutral) there will be a
coexistence between H2S and HS where hydrogen peroxide will
react with sulfides at a 1.5:1 (w/w) ratio. At high pH values (alka-
line) it will take four times as much hydrogen peroxide to turn the
S2 ion into sulfate. The reactions that occur in each pH range are
summarized in reactions (4)e(7) (Ahmad, 2010):
Acid pH:
H2 S þ H2 O2 /S þ 2H2 O (4)
Neutral pH:
Hþ þ HS þ H2 O2 /S þ 2H2 O (5)
HS þ 4H2 O2 /SO2
4 þ 4H2 O þ H
þ
(6)
Alkaline pH:
S2 þ 4H2 O2 /SO2
4 þ 4H2 O (7)
Therefore, in order to use lesser amounts of hydrogen peroxide
spent caustic should be in the acidic range where sulfides are
mainly converted to hydrogen sulfide and the oxidant will com-
plete the oxidation of any non-converted sulfides.
After neutralization coupled with classical oxidation the spent
caustic met requirements related to total sulfides were the con-
centration went down to zero whilst the concentration of COD was
still higher than recommended (1000 mg/L). Therefore, it was
decided that classical oxidation is not enough. Advanced oxidation
was performed.
3.3. Neutralization coupled with advanced oxidation
Oxidation by H2O2 alone is not effective for high concentrations
of contaminants (Huang and Tang, 1993). Therefore, a catalyst could
be used to accelerate the reaction rate of H2O2 to form more hy-
droxyl radicals. In this study iron was used as a catalyst. A solution
of hydrogen peroxide and an iron catalyst is referred to as Fenton
reagent.
108 A. Hawari et al. / Journal of Environmental Management 151 (2015) 105e112

Table 2 Table 2 (continued )

Advantages and disadvantages of the most commonly used classical chemical
oxidants. Chemical Advantages Disadvantages Ref.
oxidant
Chemical Advantages Disadvantages Ref.
 Applicable to a  Generate bromate
oxidant
wide range of in bromide-rich
Chlorine  Strong oxidant.  Generates halo- (Rodríguez, 2003) organic materials. waters.
 Strong disinfec- genated DBPs  Very strong  Can react with a
tant generate  Possibly disinfectant. variety of con-
persistent deposit. contribute to odor  Efficient for taste taminants, but it
 Cheap oxidant. and taste issues and odor control. also can react
 Very simple to  Requires high  Does not generate with many other
injected into the dosage of chlorine halogenated DBPs molecules.
system.  Could create a except in  Low solubility of
 Available in solid, carcinogenic bromide-rich ozone in aqueous
liquid or gaseous organochloride by waters. solutions.
form. products  May be used to  Lack of studies on
Chlorine  Strong oxidant  Produces chlorite (Rodríguez, 2003) support in the large scale
dioxide  Strong as an inorganic coagulation/floc- operations.
disinfectant DBP culation process.
 Gaseous (very  It might create Hydrogen  One of the cheap-  The oxidation (Gogate and Pandit,
soluble in water). unwanted odors Peroxide est oxidizers. process doesn't 2004; Stasinakis,
 Used efficiently  hard to maintain a  It has a high produce by 2008; Tsitonaki
for taste and odor persistent oxidizing product. et al., 2010)
control for specific disinfection potential.  Not effective for
types  Difficult to handle  It is water-soluble. complex
 Does not react and transport.  It does not pro- materials.
with ammonia. duce toxins or co-  Mass transfer
Oxygen  Easy to feed in the  Quite weak (Rodríguez, 2003) lor byproducts. limitations be-
system oxidant for the  Applicable for a tween the
 Does not generate majority of water wide range of hydrogen
halogenated by- treatment system. organic peroxide with the
products.  Requires certain contaminants. organic materials.
 Low operation operational  Can be combined  Over-oxidation
costs conditions. with ozone or UV reactions could
 High installation to increase the occur.
cost. efficiency.  Long residence
Persulfate  Highly stable.  New technology (Gogate and Pandit,  It can be stored time.
 High oxidation still requires more 2004; Stasinakis, and transported
potential. field studies. 2008) safely.
 Applicable to wide  Might degrade
range of organics. soft metals.
 It can be catalyzed  Undesirable long
by heat, UV light, lasting sulfate 3.3.1. Effect of pH in Fenton's reagent on COD removal
high pH, hydrogen (SO24 ) by- It can be seen from Fig. 3 that COD removal percentages were
peroxide, and products. higher than those in the case of neutralization alone and neutrali-
transition metals.
zation coupled with classical oxidation. Also in the case of advanced
 Highly reactive at
pH < 3 and at oxidation it can be seen that at acidic conditions with a pH value
pH > 10. below 4.5 a higher COD removal was accomplished. This could be
 Fewer mass due the overall Fenton chemistry which could be described by
transfer and mass
reaction (8) (Walling, 1975):
transport
limitations.
Permanganate  Easy to feed in the  Produces manga- (Rodríguez, 2003; 2Feþ2 þ H2 O2 þ 2Hþ /2Feþ3 þ 2H2 O (8)
system. nese dioxide by Gogate and Pandit,
 Does not generate product that 2004; Stasinakis, Reaction (8) shows that in order to generate maximum quan-
halogenated DBPs. should be 2008) tities of hydroxyl radicals acidic conditions are necessary. This is
 Efficient for removed. why at lower pH values higher removal percentages of COD were
certain types of  Can result in a accomplished. Also once the hydroxyl radicals oxidize the organic
odor and taste. pink water color if
 Applicable over a dosage not
compounds (RH) it will produce organic radicals (R), which are
wide pH range. controlled. very reactive and can be further oxidized as described in reaction
 Widely available  Limited disinfec- (9) (Walling and Kato, 1971; Lin and Lo, 1997):
and cheap. tion capabilities.
 Easy transport and  Reduction in effi- RH þ OH$/H2 O þ R$/ðchain propagationÞ (9)
handling. ciency due to side
 Less health and reactions.
safety problems.  Low oxidation
Where R refers to any organic contaminant, R refers to organic free
(no gas/heat potential. radicals
production).  Not effective for a This reaction is a chain propagation initiating a radical chain
wide range of oxidation which can be further oxidized according to reactions (10)
contaminants.
and (11) (Walling, 1975; Kochany et al., 1995):
Ozone  It is a very  Does not create a (Rodríguez, 2003;
powerful persistent disin- Gogate and Pandit,
oxidizing agent fectant residual. 2004; Stasinakis,
R$þH2 O2 /ROH þ OH$ (10)
 Quiet costly. 2008; Tsitonaki
et al., 2010) R$þO2 /ROO$ (11)
 A. Hawari et al. / Journal of Environmental Management 151 (2015) 105e112
Table 3
Advantages and disadvantages of different chemical oxidants used in the advanced
chemical oxidation process.
Chemical Advantages Disadvantages
oxidant
UV/O3 1. Supplemented 1. Energy and cost
disinfectant. intensive process.
2. More effective than 2. Potential for bromate
O3 or UV alone. formation but it can
3. More efficient in be controlled through
generating OH radi- pH adjustments.
cals than H2O2 & UV 3. Turbidity can inter-
for equal oxidant fere with UV light.
concentrations. 4. Ozone diffusion can
result in mass trans-
fer limitations.
5. May require ozone off
gas treatment.
UV/H2O2 1. No sludge 1. It cannot utilize solar
generation. light as the source of
2. UV/H2O2 process is UV light.
efficient in mineral- 2. H2O2 has poor UV
izing organic absorption
pollutants. characteristics.
3. No potential for 3. Turbidity can inter-
promate formation. fere with UV light.
4. No off gas treatment 4. Less stoichiometric
is required. efficiency in gener-
5. Not limited by mass ating OH radicals
transfer relative to when compared the
O3 processes. O3/H2O2 process.
5. Interference of other
compounds like ni-
trate with the UV
light.
UV/TiO2 1. Chemical stability of 1. Impossible to achieve
TiO2 in aqueous uniform radiation of
media and high po- the entire catalyst
tential to produce surface.
radicals. 2. Pretreatment is
2. Possible use of solar essential to avoid
radiation. fouling of the TiO2
3. TiO2 is a cheap, catalyst.
readily available 3. If TiO2 is added as
material. slurry then a separa-
4. TiO2 is capable to tion step is required.
oxidize a wide 4. Reaction efficiency is
range of organic highly dependent on
compounds. pH because of that
5. No potential for close monitoring and
bromate formation. control is required.
6. Can be performed at 5. Requires an onsite
high UV storage or regenera-
wavelengths. tion method.
7. Off gas treatment is
not required.
Perozone 1. Peroxone process is 1. Very dangerous and
(O3 þ H2O2 much rapid than should be carefully
using O3 or H2O2 handled when used
alone. and stored.
2. It is extremely effi- 2. Not very efficient
cient to treat com- when it is used to
plex compounds. oxidize iron and
3. H2O2 is stable in manganese.
acidic media. 3. Potential to produce
byproducts such as
aldehydes, ketones,
peroxides.
4. May require treat-
ment of excess H2O2.
5. May require gas
treatment.
Fenton 1. Fenton may lead to 1. Needs more pH
Oxidation a complete destruc- control.
H2O2 and tion of contami- 2. Hazards associated
ferrous nants under ideal with using H2O2.
iron (Fe(II)) conditions.
109
Table 3 (continued )
Chemical Advantages Disadvantages Ref.
oxidant
Ref.
2. The iron used can be 3. High concentration of
removed from the hydroxyl radicals
(Upadhye and solution. might react with
Joshi, 2012; 3. Hydrogen peroxide other unwanted
Sharma et al., is easy to store and species.
2012) to handle. 4. The reaction is
4. High oxidation po- exothermic which
tential that can could cause an in-
target complex crease in
organic compounds. temperature.
5. Treatment of both
organic and inor-
ganic substances
could be achieved.
6. Can oxidize wide
(Sharma et al., range of
2012) contaminants.
7. Iron and hydrogen
peroxide are cheap
and safe.
The organic free radicals produced in reaction (2.9) may then be
oxidized, reduced, or demineralized according to reactions
(12)e(14) (Tang and Tassos, 1997):
R$ þ Feþ3 /Rþ þ Feþ2 ðoxidationÞ (12)


R$ þ Feþ2 /R þ Feþ3 reduction (13)
(Upadhye and
Joshi, 2012;
2R$/R  R ðdimerizationÞ (14)
Sharma et al.,
2012) Fenton's reagent can accomplish both chemical oxidation as
well as coagulation treatment. As mentioned earlier the pH value
was adjusted to 7 after the reaction was accomplished. Adjusting
pH to a value around 7 after the reaction was accomplished will
result in precipitating the dissolved ferric particle. Precipitated
ferric particles will combine to form flocs that will further help in
removing other dissolved particles. As a result Fenton's method can
achieve both chemical oxidation as well as coagulation/flocculation
treatment.
3.3.2. Effect of hydrogen peroxide/ferrous sulfate ratio on COD
removal
It is critical to find the best ratio of hydrogen peroxide to ferrous
sulfate that will generate the highest amount of hydroxyl radicals.
(Upadhye and Hydroxyl radicals will oxidize contaminants in the spent caustic
Joshi, 2012)
solution which will result in a better COD removal. As COD removal
increases, more hydroxyl radicals are being generated and vise
versa (Tsitonaki et al., 2010). Fig. 4 shows the COD % removal for
different ratios of hydrogen peroxide to ferrous sulfate at a pH value
of 2.5. It can be seen from Fig. 4 that the COD removal increased as
the hydrogen peroxide/ferrous sulfate ratio increased up to a
certain point were after that point the removal percentage of COD
starts to decrease. The maximum COD removal of 96.5% was ach-
ieved at a hydrogen peroxide to ferrous sulfate ratio of (7:1).
Depending on the H2O2/FeSO4 ratio different COD removal
mechanisms could be involved. Three main mechanisms could be
identified: Physical separation is dominant when the amount of
(Upadhye and Feþ2 employed exceeds that of H2O2. Physical separation occurs by
Joshi, 2012;
coagulation and flocculation processes where ferric particles will
Sharma et al.,
2012) combine to form larger flocs that will swap other dissolved parti-
cles while precipitating. This is considered to be undesirable since
the aim of the treatment process is oxidation (Tsitonaki et al., 2010).
110 A. Hawari et al. / Journal of Environmental Management 151 (2015) 105e112

Table 4
Ethylene plant spent caustic characteristics used in this study.

Parameters Unit Standard Range (min. e

method max.)

COD mg/L HACH 14,700e15,300

Program
BOD mg/L SM 5210 B 400e437
pH 12.8e13.7
Conductivity ms/cm 125e140
Oil & grease mg/L ASTMD7678 59e72
Sulfide mg/L HACH 5050e5310
Program
Total Solids (TS) mg/L ASTMD5907 85,120e85,550
Total Dissolved Solids mg/L ASTMD5907 77,200e78,100
(TDS)
Total suspended Solids mg/L ASTMD5907 8020e8235
(TSS)
Sulfate mg/L EPA 300.1 7995e8230
Chloride mg/L EPA 300.1 11.8e15.2
Phenol mg/L EPA 420.1 0.06e0.21
Heavy metals mg/L EPA 200.7 Traces Fig. 3. Effect of pH on the COD removal percentage with advanced oxidation process
Free soda Weight percentage Manual 0.9e2.6 (AOP), neutralization, and classical oxidation.
(wt. %) titration
Total alkalinity Weight percentage Manual 3.0e3.9
(wt.%) titration
100

Chemical Oxygen Demand (COD) removal

98
96
94
percentage ( %)

92
90
88
86
84
82
80
0 200 400 600 800
Ratio of hydrogen peroxide to iron sulphate [H2O2/FeSO4]

Fig. 4. Effect of hydrogen peroxide to ferrous sulfate ratio on COD % removal.

Fig. 1. COD and sulfide removal percentage at different pH values.

Also when Feþ2 employed exceeds that of H2O2 ferrous ions will
compete with contaminants to react with hydroxyl radicals
100 100 (Neyens and Baeyens, 2003). Ferrous ions will terminate hydroxyl
Sulfide
Chemical Oxygen Deman (COD) removal

98 98 radicals as shown in reaction (15). This reaction can convert ferrous

pH = 3.5
Sulfide removal percentage ( %)

96 96 ions from a catalyst to a reactant. This leads to lower hydroxyl

pH = 2.5
radical utilization for the oxidation of contaminants.
94 pH = 1.5 94
percentage ( %)

92 92
90 90
88
COD
88 OH$þFeþ2 /OH þFeþ3 (15)
86 COD 86 The second mechanism is chemical oxidation. Chemical oxida-
84 84 tion is dominant when the amount of hydrogen peroxide is much
COD higher than ferrous ions (Neyens and Baeyens, 2003). This case is
82 82
desirable because the aim of applying Fenton's reagent is the
80 80 oxidation of contaminants rather than physical separation. How-
0 50 100 150 200
ever, the issue with this ratio is competition of hydrogen peroxide
Concentration of hydrogen peroxide (H2O2) (mM/L) and contaminants to react with the hydroxyl radicals. Hydroxyl
radicals tend to react with hydrogen peroxide instead of reacting
Fig. 2. COD and sulfide removal % at different H2O2 concentrations and at different pH
with contaminants generating HO2 as shown in reaction (18).
values.
Moreover, HO2 reacts with ferrous or ferric ions as shown in
reactions (16) and (17). This will result is loss of removal efficiency
of COD.
 A. Hawari et al. / Journal of Environmental Management 151 (2015) 105e112
Feþ2 þ HO2 $/Feþ3 þ HO2
Feþ3 þ HO2 $/Feþ2 þ O2 þHþ
OH$þH2 O2 /H2 O þ HO2 $
The third mechanism is also chemical oxidation but with a
medium ratio. This occurs when the amount of hydrogen peroxide
to ferrous ions falls in between the previous two extremes. At this
range, hydroxyl radicals tend to react with contaminants instead of
hydrogen peroxide or ferrous ions as shown in reaction (19). At this
favorable range, ferrous ions react with hydrogen peroxide rather
than reacting with hydroxyl radicals and a maximum amount of
hydroxyl radicals would be generated (Neyens and Baeyens, 2003).
As mentioned before in this study the best hydrogen peroxide to
ferrous sulfate ratio was found to be 7:1 with a 96.5% COD removal.
RH þ OH$/H2 O þ R$
3.3.3. Amount of hydrogen peroxide to initial COD concentration
The optimum ratio of hydrogen peroxide to ferrous sulfate ratio
was found in the previous section. However the amount of
hydrogen peroxide to spent caustic initial COD concentration still
needs to be determined. This amount is essential in Fenton's re-
action because it verifies the amount of hydrogen peroxide that
should be used to carry out the required treatment. The cost of
treatment by Fenton's process mainly depends on the cost of
chemicals. Chemicals used in Fenton's reaction are hydrogen
peroxide, ferrous sulfate, sulfuric acid and sodium hydroxide.
Hydrogen peroxide is considered to be the most expensive among
the other used chemicals. Therefore, the running cost of the process
will mainly depend on the hydrogen peroxide amount which will
depend on the initial COD concentration in the spent caustic so-
lution. Different experiments were performed at a constant H2O2/
FeSO4 ratio of 7.4 but different concentrations of H2O2 and FeSO4, at
a pH ¼ 2.5, and constant initial COD concentration (15,000 mg/L).
From Fig. 5 it can be seen that as the hydrogen peroxide to initial
COD ratio increases, COD removal decrease. This could be due to the
competition of hydrogen peroxide and contaminants to react with
hydroxyl radicals. Hydroxyl radicals will react with hydrogen
peroxide instead of reacting with contaminants which will
decrease the efficiency of the process as discussed previously. The
maximum COD removal percentage achieved was 96.5% with a
ratio of 1.23 H2O2/initial COD (g/g). The final COD concentration
Fig. 5. Amount of hydrogen peroxide to initial COD concentration related to COD %
removal at a constant H2O2/FeSO4 ratio ¼ 7.4.
111
obtained was around 540 mg/L which satisfies the desired con-
(16) centration of less than 1000 mg/L.
(17)
4. Conclusions
(18)
Different methods to treat spent caustic produced from an
ethylene plant have been investigated. It was found that Ethylene
plant spent caustic had a relatively low COD and BOD concentra-
tions compared to other refinery spent caustic solutions. Moreover,
sulfide and phenol concentrations were also relatively lower than
other refinery spent caustic.
Different treatment processes were tested for the treatment of
ethylene plant spent caustic:
 Neutralization:
(19) Maximum removal of sulfide was at pH values below 5.5.
Decreasing the pH value below 5.5 had a minimal effect on the
removal percentage of sulfide where at pH values of 1.5, 3.5 and 5.5
almost similar removal percentages were obtained. It was also
observed that the decrease of sulfides was accompanied by a
decrease of COD. The higher percentage removal of sulfides (99% at
pH ¼ 1.5) was accompanied with the highest COD removal (88%).
This could be due to the accompanied removal of sulfide COD. The
remaining COD is due to the remaining non-sulfide compounds
which could be mainly due to emulsified hydrocarbons.
 Neutralization coupled with classical oxidation:
The maximum COD removal percentage reached 89% at pH ¼ 2.5
and at a hydrogen peroxide concentration of 19 mM/L. Above this
concentration COD removal started to decrease as the concentra-
tion of hydrogen peroxide increased. The COD removal increases as
the concentration of hydrogen peroxide increases due to the gen-
eration of hydroxyl radicals. Hydroxyl radicals tend to further
oxidize existing organic pollutants. Even at a very low concentra-
tions of H2O2 (2 mM/L) sulfides were totally removed from solution.
 Neutralization coupled with advanced oxidation using Fenton
reagent:
 At pH ¼ 2.5, a maximum COD removal of 96.4% was achieved
at a hydrogen peroxide to ferrous sulfate ratio of (7:1).
 As the hydrogen peroxide to COD ratio increases, COD removal
decrease. This is due to the competition of hydrogen peroxide
and contaminants to react with hydroxyl radicals. The
maximum tested COD removal achieved was 96.4% with a
ratio 1.23 H2O2/COD (g/g) at pH ¼ 2.5.
In order to reach the discharge limit of COD less than 1000 mg/L
and sulfide 2 mg/L it is recommended to use Fenton's reagent. The
lowest COD concentration obtained was 542 mg/L at the following
treatment conditions: a pH value of 2.5, 7.4 FeSO4/H2O2 Ratio, and
1.23 H2O2/COD (g/g).
Acknowledgments
The authors wish to acknowledge Qatar University for the
financial support through an internal student grant (QUST-CENG-
SPR-12/13-17). Also the authors would like to thank Dr. Ahmed Al
Khatat for the help provided in the chemical analysis. Moreover, the
authors would like to thank Qatar Petrochemical Company
(QAPCO) for the supply of samples.
112 A. Hawari et al. / Journal of Environmental Management 151 (2015) 105e112

References Munter, R., 2001. Advanced oxidation processesecurrent status and prospects.
Chemistry 50, 59e80.
Neyens, E., Baeyens, J., 2003. A review of classic fenton's peroxidation as an
Ahmad, W., 2010. Neutralization of Spent Caustic from LPG Plant at Preem AB
advanced oxidation technique. J. Hazard. Mater. 98, 33e50.
Goteborg. A Master thesis at the department of chemical and biological engi-
€ teborg, Sweden. Olmos, A., Olguin, P., Fajardo, C., Razo, E., Monroy, O., 2004. Physical characterization
neering. Chalmers University of Technology, Go
of spent caustic from the OXIMR process and sour water from Mexican oil re-
Alnaizy, R., 2008. Economic analysis for wet oxidation processes for the treatment
fineries. Energy Fuel 18, 302e304.
of mixed refinery spent caustic. Environ. Prog. Sustain. Energy 27, 295e351.
Rodríguez, M., 2003. Fenton and UV-vis Based Advanced Oxidation Processes in
Canizares, P., Paz, R., Saez, C., Rodrigo, M., 2009. Costs of the electrochemical
Wastewater Treatment: Degradation, Mineralization and Biodegradability
oxidation of wastewaters: a comparison with ozonation and fenton oxidation
Enhancement. A Ph.D. thesis at the Department of chemical engineering. Bar-
processes. Environ. Manag. 90, 410e420.
celona University, Spain.
Gogate, P., Pandit, A., 2004. A review of imperative technologies for wastewater
Sharma, V., Triantis, T., Antoniou, M., Pelaez, X., Dionysiou, D., 2012. Destruction of
treatment I: oxidation technologies at ambient conditions. India Adv. Environ.
microcystins by conventional and advanced oxidation processes: a review. Sep.
Res. 8, 501e551.
Purif. Technol. 91, 3e17.
Hashemi, S., Heidarinasab, A., 2012. Spent caustic bioregeneration by using Thio-
Sheu, S., Weng, H., 2001. Treatment of olefin plant spent caustic by combination of
bacillus denitrificans bacteria. World Acad. Sci. Eng. Technol. 67, 226e228.
neutralization and fenton reaction. Water Res. 35, 2017e2021.
Huang, C., Tang, C., 1993. Advanced chemical oxidation: its present role and po-
Stasinakis, A., 2008. Use of selected advanced oxidation processes (AOPs) for
tential future in hazardous waste treatment. Waste Manag. 13, 361e377.
wastewater treatment: a mini review. Glob. Nest J. 10, 376e385.
Kang, Y., Cho, M., Hwang, K., 1999. Correction of hydrogen peroxide interference on
Tang, W., Tassos, S., 1997. Oxidation kinetics and mechanisms of trihalomethanes by
standard chemical oxygen demand test. Water Res. 33, 1247e1251.
fenton's reagent. Water Res. 5, 1117e1125.
Kochany, E., Sprah, G., Harms, S., 1995. Influence of some groundwater and surface
Tsitonaki, A., Petri, B., Crimi, M., Mosbaek, H., Siegrist, L., Bjerg, L., 2010. In situ
waters constituents on the degradation of 4-chlorophenol by the fenton reac-
chemical oxidation of contaminated soil and groundwater using persulfate: a
tion. Chemosphere 9, 9e20.
review. Environ. Sci. Technol. 40, 55e91.
Kumfer, B., Felch, C., Maugans, C., 2010. Wet air oxidation treatment of spent caustic
Upadhye, V., Joshi, S., 2012. Advances in wastewater treatment- a review. Chem. Sci.
in petroleum refineries. In: National Petroleum Refiner's Association Confer-
Appl. 3, 264e326.
ence, Phoenix, AZ, pp. 21e23.
Walling, C., 1975. Fenton's reagent revisited. Acc. Chem. Res. 8, 125e131.
Lin, S., Lo, C., 1997. Fenton process for treatment of desizing wastewater. Water Res.
Walling, C., Kato, S., 1971. The oxidation of alcohols by Fenton's reagent: the effect of
8, 2050e2056.
copper ion. J. Am. Chem. Soc. 93, 4275e4281.
Miessler, G., 1991. Inorganic Chemistry, second ed. Prentice Hall (Chapter 6): Acid-
Base and Donor-Acceptor Chemistry.