Fuel xxx (xxxx) xxxx

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Effect of steam jet on oil reclamation and purification from layered oily
sludge
Dong Xiaa,b,c, Kai Hea,b, Zhi-Yuan Zhanga, Fu-Shen Zhanga,b,
⁎

a
Department of Solid Waste Treatment and Recycling, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China
b
University of Chinese Academy of Sciences, Beijing 100049, China
c
Energy Research Institute @ NTU (ERI@N), Nanyang Technological University, 637553 Singapore, Singapore

GRAPHICAL ABSTRACT

ARTICLE INFO ABSTRACT

Keywords: Steam is widely used to enhance heavy oil recovery, whereas it is implemented as just an auxiliary thermal source
Oily sludge when treating oily sludge. In this study, steam jet was for the first time introduced to simultaneously reclaim and
Steam jet purify oil from layered oily sludge, a hazardous waste in storage ponds of oil joint stations. The effect of steam jet on
Oil reclamation oil reclamation and purification was investigated with various operating temperatures (200–350 °C) and mass ratios
Demetallization
of steam to oily sludge (2:1–8:1). Batch experiments indicated that a high steam jet performance was achieved at
Numerical simulation
300 °C and 6:1 of the mass ratio, with 92% of oil reclamation rate and 82% of demetallization rate from the oily
sludge of upper layer within 5 min. The mild temperature of the injection area (97 °C) was in favor of retaining light
fractions of oil. Besides, the dynamic process of this steam-based treatment was numerically simulated with COMSOL
to illuminate interactions among steam, oil and condensed water. With a significant reduction of oil viscosity due to
steam heating, high dynamic pressure of steam jet was found to disturb the surface of layered oily sludge, expediting
oil flotation in water. Meanwhile, the intense steam jet could also sufficiently scatter the floated oil to diffuse
metallic impurities from oil droplets into surrounding water. The reclaimed oil after steam injection was more
upgradable for a refinery while the residual oily sludge became more treatable with traditional recycling.

⁎
Corresponding author at: Department of Solid Waste Treatment and Recycling, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences,
Beijing 100085, China.
E-mail address: fszhang@rcees.ac.cn (F.-S. Zhang).

https://doi.org/10.1016/j.fuel.2019.116731
Received 20 August 2019; Received in revised form 28 October 2019; Accepted 21 November 2019
0016-2361/ © 2019 Elsevier Ltd. All rights reserved.

Please cite this article as: Dong Xia, et al., Fuel, https://doi.org/10.1016/j.fuel.2019.116731
D. Xia, et al. Fuel xxx (xxxx) xxxx

1. Introduction drainage (SAGD) and cyclic steam stimulation (CSS), are extensively
used to enhance recovery of heavy oil [30]. To reduce oil viscosity and
Oily sludge is inevitably generated during each process of crude improve oil mobility, steam is typically injected with a temperature
exploitation, storage, transportation, and refining, leading to a world- ranging from 240 to 350 °C and a corresponding pressure ranging from
wide enormous accumulation (over 1 billion tons) [1,2]. As oily sludge 2.0 to 19.5 MPa [31,32]. Although steam offers several advantages in
contains a high concentration of petroleum hydrocarbons (PHCs), terms of comprehensive facilities and sustainable steam source [33,34],
heavy metals and other toxic components, it is regulated as a hazardous there are few reports about adopting this technique to reclaim oil from
waste in many countries [3]. In recent years, the intense demand for oily sludge directly but just as an auxiliary heat source [10,35,36].
crude oil coupled with the declining conventional oil reserves has re- McCoy et al. [37] disclosed a patent employing a continuous process of
sulted in an urgent need of unconventional substitutes, such as heavy steam distillation to deoil and dewater refinery sludge at a temperature
and extra-heavy oils, which accounts for two-thirds of total world oil varying from 150 to 315 °C. However, some heavy oil fractions could
reserves [4,5]. The exploitation of a considerable volume of heavy oil not be vaporized due to their high boiling points. Cheng et al. [38]
greatly enhances the challenge of harnessing oily sludge due to its high combined steam injection with the pyrolysis of oily sludge to improve
viscosity, acidity and asphaltene contents [6,7]. the yield and quality of oil products, but this method is difficult to
Many efforts have been devoted to treating oily sludge. The re- realize the full utilization of sufficient steam. Accordingly, a more cost-
clamation of valuable oil from oily sludge is considered as a primary effective and efficient method based on the clean steam source to treat
purpose due to its substantial financial and environmental benefits oily sludge could be a significant technique improvement.
[2,8]. Various methods are available for oil reclamation, including In this study, a novel method taking full advantage of sustainable
solvent extraction [9–11], thermochemical washing [12], centrifuga- and sufficient steam sources was developed to process layered oily
tion [13], freeze/thaw [14–16], ultrasonic irradiation [17], froth flo- sludge of storage ponds. When steam jet was vertically injected onto the
tation [18], and pyrolysis [19–21]. However, each method possesses layered oily sludge, lighter oily sludge from the upper layer could be
different treatment capacity, efficiency, and costs, which limit their separated rapidly and efficiently from the bottom one. In the meantime,
applicability. Moreover, the components of oily sludge vary over a wide the floated oil was continuously injected by the steam to decontaminate
range depending on their sources of generation. Even in the same sto- metallic impurities, thus realizing a concurrent effect of oil reclamation
rage pond, oily sludge could be heterogeneous due to various proces- and purification. Therefore, it is promising to utilize clean steam jet to
sing schemes and the precipitation of heavier compositions during alleviate the burden of oily sludge storage as well as reclaim oil with
storage, leading to stratification [22,23]. Therefore, none of any single higher refineability in heavy oil production areas.
method currently used is sufficient to treat all types of oily sludge.
The quality of reclaimed oil is of another concern since it would
ultimately be redirected to a refinery for upgrading as a substitute of 2. Materials and methods
raw petroleum [10]. Compared with raw petroleum, reclaimed oil may
contain more metallic impurities, which may cause installation corro- 2.1. Materials
sion and catalyst deactivation during refining [24,25]. It is therefore
essential to remove the metallic impurities to the highest achievable Surface oily sludge (S-OS), a highly viscous liquid, as well as bottom
quality. Among current methods, ultrasonic irradiation and cen- oily sludge (B-OS), a semi-solid substance, was collected at different
trifugation demonstrate a desirable performance in removing water- depths from a typical concrete-lined storage pond located in a crude
soluble salts [26], while solvent extraction can remove oil-soluble joint station of Shengli Oil Field, China. Samples were sealed in sepa-
metals (e.g. Ni and V) [27]. In addition, supercritical water and me- rated buckets at ambient temperature. Each sample was well mixed
thanol can separate Ni, V, Fe and Ca from petroleum emulsion [28,29]. before experiments. Analytical grade chemicals including toluene, n-
However, these methods consume a large amount of solvent or require hexane, dichloromethane, ethyl alcohol, trichloromethane, nitric acid,
complicated equipment. perchloric acid, and hydrofluoric acid were purchased from Sinopharm
Steam-based thermal approaches, such as steam-assisted gravity Chemical Reagent Shanghai Co., Ltd.

Fig. 1. Steam jet generator for processing oily sludge.

2
D. Xia, et al. Fuel xxx (xxxx) xxxx

2.2. Equipment of steam jet treatment separated by precipitation of asphaltenes with n-hexane, followed by
further separation through a chromatographic column, and were
An electrically heated kettle (HAFA-2, Huaan Co., Ltd., China) weighted based on ASTM D2007-11. Elements (C, H, and N) were
coupled with a high-pressure Hadtelloy alloy container (100 mL) was measured using an elemental analyzer (EA3000, EuroVector). The
used to generate steam, as demonstrated in Fig. 1. The steam was in- sulfur content was measured with an automatic coulomb sulfur ana-
jected into a glass tube through a steam line of 2 mm in diameter lyzer (5E-AS3200B, CKIC).
covered with cotton cozy. A perforated plug was mounted on the top of Multi-metal concentration was determined with Inductively
the glass tube to fix the steam line and release excess steam. A straight Coupled Plasma Atomic Emission Spectrometry (ICP-AES, Prodigy7,
pipe nozzle was located at the end of the steam line as the outlet of the Leemans) after digestion in mixed acids.
steam and was fixed at a height of 6 cm above the surface level of oily Differential Scanning Calorimetry (DSC) instead of ASTM D 95-13
sludge for each experiment. To acquire the steam properties, an inner was employed to detect emulsified water retained in the reclaimed oil
thermocouple and a pressure transducer were used to detect the tem- because it is more accurate for a small amount of water in the sample
perature and pressure of the steam in the container. Another thermo- [39]. The measurement of thermogram was performed in a DSC
couple (OM-EL-USB-TC-LCD, Omega) was inserted into the tube to (Q2000, TA Instruments), and the total water content w (%) can be
monitor the temperature of the injection area and deduce the total in- calculated as follows [39]:
jection time.
w = Q w / Hf (2)

2.3. Evaluation of steam jet performance tb

Qw = Q·dt
ta (3)
The performance of steam jet was judged by its effect on oil re-
where Qw is the specific heat absorbed during melting, J/g; Hf is the
clamation and purification. The efficiency of oil reclamation was
fusion enthalpy of frozen water and its value is 233 J/g; Q is the heat
measured as an oil reclamation rate (ORR, %). A certain volume of
flux (W/g); ta and tb are the beginning and the ending time of melting
deionized water (20, 40, 60 and 80 mL) was sealed in the container and
during DSC analysis.
was heated to a set temperature (200, 250, 300 and 350 °C). 10 g of S-
In addition, microstructures of the oily sludge and the reclaimed oil
OS was squeezed into a glass tube by a syringe, thus the mass ratio of
were observed using an optical microscope (BA210, Motic) with 100
steam to oily sludge (S/OS) ranged from 2:1 to 8:1. After the steam
times magnification.
injection, three phases were formed (floated oil emulsion, intermediate
water and oily sludge remnants from the top-down). The sample was
immediately cooled down in an ice-water bath to minimize natural 2.5. Numerical simulation
flotation. Meanwhile, 40 mL of toluene was added into the tube and
was shaken at 150 rpm for 5 min to dilute floated oil emulsion. Hereby, COMSOL is a commercially available software, reliable to simulate
the whole oil–toluene mixture from the top layer was easily transferred in heat transfer, electromagnetism, structural mechanics and fluid dy-
into a separation funnel for further dewatering. Finally, the amount of namics [40,41]. In this study, the fluid dynamics associated with the
oil was weighted with an analytic balance (AL104, Mettler Toledo) after steam jet, the oil reclamation and purification processes were simulated
removing toluene in a vented oven at 50 °C until constant weight. The by COMSOL Multiphysics 5.3 with a CFD Module, and the steam
ORR was defined as: properties were calculated by SteamTab software.
The simulation of the steam jet with different operating tempera-
Mr ·(1 w ) tures was implemented by a k–ω Turbulent Model because of its high
ORR = × 100%
Mos·Oos (1) Reynolds numbers [42]. According to Bernoulli's equation, dynamic
where Mos and Mr indicate the weight of oily sludge and reclaimed oil pressure, along with static pressure and pressure due to elevation, is
respectively; Oos is the percentage of oil content of the oily sludge; w derived from the conversion of energy for a fluid in motion. In a sim-
denotes the percentage of water content of the reclaimed oil. plified case, when the steam jet is stagnated due to impact, its fluid
The purification performance of steam jet was evaluated as de- velocity turns to zero and all kinetic energy is converted into pressure
scribed [13]. Briefly, after steam injection, the floated emulsion was energy. The dynamic pressure (q) is the kinetic energy per unit volume
promptly transferred into a centrifuge tube without cooling or adding of steam and is defined as [43]:
toluene and was heated to 80 °C prior to centrifuge at 3000 rpm for q = 1/2 u2
30 min. After the centrifugation, the water content, multi-metal con-
centration, and chemical composition of the upper layer of the re- where ρ and u are the density and the velocity of the steam. The profile
claimed oil were analyzed (see Section 2.4). Each experimental condi- distribution of dynamic pressure was used to describe the flow property
tion was set in duplicate, and the result was shown as the mean value. of the steam jet.
For the oil reclamation process, the simulation was based on time-
2.4. Characterization of oily sludge and reclaimed oil dependent analyses of models in which Laminar Two-Phase Flow and
Level Set Interface were used [44]. Besides, the oil purification process
The oily sludge samples were put into a filter cartridge for azeo- was modeled with the turbulence two-phase flow (k–ε) and Level Set
tropic distillation with toluene, by which the contents of oil, water and Interface. All the equations of models and boundaries are defined in the
solids were separated and measured following ASTM D 95-13. The CFD module and are presented in Appendix A.
extracted oil of S-OS and B-OS, denoted by S-EO and B-EO respectively,
were obtained after removing toluene as described in Section 2.3. 3. Results and discussion
The densities of oily sludge and oil were determined by pycnometry
at 20 °C. Viscosity (μ) measurements were performed in a digital ro- 3.1. Characteristics of the oily sludge
tational viscometer (CAP2000+H, Brookfield) with spindle No. 04 at a
constant shearing rate of 50 rpm from 50 to 105 °C. Higher heating Due to the distinct layering phenomenon in the storage pond over
value (HHV) was measured with an oxygen-bomb calorimeter several years of precipitation, S-OS and B-OS presented an obvious
(SDACM3100, Sundy) according to ASTM D240-17. difference in appearance. From a microscopic perspective, spherical
Chemical content of oil phase was analyzed as following: SARA water droplets of diameters less than 30 μm were well dispersed in the
components (saturates, aromatics, resins, and asphaltenes) were first oil phase of S-OS, while irregular solid particles (black and bright) of

3
D. Xia, et al. Fuel xxx (xxxx) xxxx

different sizes were distributed in the oil phase of B-OS (Fig. B.1). The was even more significant in its oil phase (S-EO) (Fig. 3). The data could
differences in the makeup of the three-phase components of these two be concluded that increasing temperature could greatly ameliorate the
samples (Table 1) may explain the differences observed in the micro- oil’s mobility and ease its separation, consistent with previous studies
structure. Oil accounted for about 93% in S-OS, with water (6.56%) as [34,46].
the second major component, while the remaining was a tiny amount of However, the temperature of the injection area was constant at
solids. By contrast, B-OS contained more water and solid impurities but around 97 °C during the stable stage for all batch experiments (Fig.
less oil than S-OS. As a consequence, S-OS possessed a lower density B.3). It could not be alone to affect oil reclamation performance
and viscosity but a higher HHV compared to B-OS. Notably, both S-OS (Fig. 2a). Therefore, the flow property of steam might also play a role.
and B-OS were too viscous to be pumped even at 50 °C, although their The dynamic pressure profiles of the steam at different temperatures
densities were on either side of 1 g/cm3. The water contents determined were simulated with COMSOL and were shown in Fig. 4. The actual
by DSC were consistent with the data determined by the standard steam temperature input into the software was based on the steam
method ASTM D 95-13, which agreed with the conclusion in a previous temperature at the beginning of the stable injection stage. In each
study [39]. profile, the farther away from the nozzle, the lower the dynamic pres-
The composition and properties of S-EO and B-EO were listed in sure. Because the oil level raised with steam processing, it was hard to
Table 2. S-EO contained more saturates and fewer asphaltenes than B- predict the exact dynamic pressure working on the oil level. But no-
EO. Since heteroatoms are enriched more in asphaltenes [45], a higher tably, steam possessed higher dynamic pressure as the temperature was
content of N was presented in B-EO. Similar to the layered oil sludge, S- increased, thus creating a stronger impact. Therefore, the steam jet
EO had a slightly lower density and viscosity but a higher HHV com- provided not only a heat source but also high kinetic energy, which
pared to B-EO. Based on the fact that the total sulfur levels in both resulted in the desired reclamation performance.
samples were well above 0.5%, these oils belonged to sour crude oil. In In addition, with the separation of S-OS, the interface of B-OS might
addition, the densities of both oils ranged from 0.94 to 0.95 g/cm3 be gradually exposed. For investigating the effect of B-OS on oil re-
(corresponding API was 18–19), and their viscosities were far beyond clamation, 5 g B-OS, 5 g S-OS and their layered combination were se-
0.1 Pa∙s, which put both the oils to the heavy crude oil category [34]. parately loaded. The operating condition was set at 300 °C with an S/
The multi-metal concentrations of the oily sludge and its extracted oil OS of 6:1. As shown in Fig. B.4, ORR reached 55% within 80 s in S-OS
were shown in Table B.1. Most of the metals in S-EO were less than while barely any ORR could be obtained from B-OS, and it was almost
those in B-EO, except for Ni, which was almost the same for both layers reduced by half when B-OS underlay S-OS. This data provided con-
and accumulated in the oil phase. Similarly, metal contents in S-OS vincing evidence that B-OS was not able to be reclaimed with the steam
were much less than those in B-OS. jet. Meanwhile, it had a hindering effect on oil recovery near the
The higher ratio of oil in S-OS with fewer impurities and heteroa- boundary of two layers because of its high viscosity and density.
toms made it highly potential for direct oil reclamation, while the re- Therefore, a low density range of oily sludge (ρ < 1 g/cm3) was a
sidual B-OS with lower content of oil was more suitable for traditional prerequisite for oil reclamation when using the steam jet.
treatments [9–12]. Such layered oily sludge was common among over
fifty joint stations of Shengli Oil Field. Recycling of oil from the oily
sludge from the upper layer not only could be achievable but also sig- 3.3. Effect of steam jet on oil purification
nificant for alleviating high storage burden for joint stations in heavy
oil production areas. Besides oil reclamation, the effect of the steam jet on oil purification
was also investigated, including the removal of metallic impurities and
3.2. Effect of steam jet on oil reclamation the changes in chemical compositions of the oil product. The experi-
mental conditions were set at 300 °C and 350 °C with an S/OS of 6:1
To establish the operating conditions with the steam jet, steam and designated as SI-300-6 and SI-350-6 respectively. The blank con-
temperature and S/OS were investigated for their effects on the per- trol, S-OS-C, was heated and centrifuged without steam jet treatment.
formance of oil reclamation from S-OS. Generally, water was trans- The multi-metal concentrations were measured in samples before
formed into saturated steam except for the loading of 80 mL water and after oil reclamation and data were shown in Fig. 5. With the set-
beyond 300 °C, which displayed in a state of subcritical water owing to tling behaviors of dispersed water and solids during centrifugation,
the extra loading of water (Fig. B.2). According to the changes in most of the metal elements in S-OS-C were reduced including Fe, Mg, K,
temperature and pressure, three distinct stages can be identified in one Ca, Al, and Na compared with those in the original S-OS. However, the
batch of the steam jet process, namely, tube-warming, stable injection concentrations of Ni and V were found to be similar or higher after
and the end of injection (Fig. B.3). reclamation, most likely due to their embedment in asphaltenes as
As shown in Fig. 2a, higher ORR was obtained with the increase of metal porphyrin structures [24,28]. When the steam injection was
temperature and S/OS. When the steam was heated at 300 °C and S/OS performed at either condition, it could further enhance the demetalli-
was set at 6:1, ORR was as high as 92%. Further increase of temperature zation effect on the reclaimed oil. The overall demetallization perfor-
to 350 °C and increase of S/OS to 8:1 could ultimately enhance ORR to mance was better at 300 °C than at 350 °C, which could be owing to a
99%, it would inevitably increase capital and operating costs due to more lasting steam jet (Fig. 2b). Total metal contents were reduced by
ultra-high pressure and extra water consumption. On the other hand, 82% in SI-300-6, among which the water-soluble light metals were
temperature and S/OS had an opposite effect on the time of duration for
the steam injection process (Fig. 2b). The elevated temperature shor-
Table 1
tened the time, while the increased S/OS prolonged the time. Therefore, Composition and physical properties of the oily sludge.
it would be more efficient to operate at a high temperature and a low S/
OS. Taken both the oil reclamation performance and economic cost into S-OS B-OS

consideration, the following operating conditions were conducted at Oil (wt%) 93.02 ± 0.94 37.59 ± 3.82
300 °C and 350 °C with an S/OS of 6:1. Water (ASTM D 95-13, wt%) 6.56 ± 1.20 21.88 ± 2.14
To understand the influence of operating temperature on oil re- Solid (wt%) 0.42 ± 0.50 40.53 ± 3.46
clamation, the viscosities of the oily sludge and the flow properties of Water (DSC, wt%) 7.22 ± 0.47 21.94 ± 0.44
Density (g/cm3) 0.9423 ± 0.0171 1.2346 ± 0.0104
the steam were investigated. As shown in Fig. 3, while the bulk S-OS
Viscosity (Pa∙s, 50 °C) 2.4 ± 0.2 16.5 ± 1.4
was heated from 50 to 100 °C, its viscosity decreased from 2.4 to HHV (J/g) 37470 ± 107 14980 ± 200
0.03 Pa∙s. This phenomenon of viscosity drop (from 4.5 to 0.11 Pa∙s)

4
D. Xia, et al. Fuel xxx (xxxx) xxxx

Table 2
SARA composition and physical properties of the extracted oil from the oily
sludge.
S-EO B-EO

SARA (wt%)
Saturates 33.60 ± 1.94 26.90 ± 2.14
Aromatics 30.36 ± 1.17 33.53 ± 1.77
Resins 23.89 ± 0.41 22.03 ± 1.08
Asphaltenes 8.91 ± 0.36 13.65 ± 0.67

Ultimate analysis (wt%)

C 86.17 ± 0.12 84.98 ± 0.14
H 11.09 ± 0.23 11.18 ± 0.09
N 0.99 ± 0.09 1.11 ± 0.01
S 2.67 ± 0.01 2.64 ± 0.00

Density (g/cm3) 0.9380 ± 0.0003 0.9462 ± 0.0057

Viscosity (Pa∙s, 50 °C) 4.5 ± 0.4 6.9 ± 0.6
HHV (J/g) 41920 ± 8 41281 ± 204

Fig. 3. Viscosity changes of the oily sludge and corresponding extracted oil
with temperature.
removed by 91% and Fe and Zn were removed by 41% and 17% re-
spectively. Interestingly, V was reduced to under detection limit and Ni
was also reduced by 5% after the steam injection, indicating a possible 3.4. Mechanism study on oil reclamation and purification
removal of some porphyrin structure, although the temperature of the
injection area remained at only 97 °C. Different mechanisms of metal To investigate the mechanisms of the steam jet on oil reclamation
removal is discussed in Section 3.4. and purification from oily sludge, the dynamic processes were sepa-
During the steam jet treatment, condensed steam could inevitably rated into two steps and were simulated with COMSOL Multiphysics
introduce water into the oil product. The water contents of S-OS-C, SI- respectively.
300-6, and SI-350-6 deduced from DSC were 4.8%, 8.5%, and 8.7% First, the oil reclamation process can be described as four stages
respectively (Fig. B.5). Microscopic images showed that the diameters according to the numerical simulation (Fig. 7): (I) order-organized oil
of these water droplets were less than 10 μm (Fig. B.6), indicating that layer was disturbed by steam jet with its viscosity reduction; (II) oil–-
the steam jet created more emulsified water which was hard to be re- water contact line shrank to form oil droplets owing to the penetration
moved by centrifugation [47]. Therefore, deep dewatering is required of water into oil phase [52]; (III) the oil droplets became elongated and
for the reclaimed oil through electrical dehydration or chemical de- detached from the surface oily sludge under buoyancy [53]; (IV) water
mulsification [48–50]. channels were formed, and the oil floated along a wall to the top. The
The chemical composition of the reclaimed oil after the steam jet numerical simulation exhibited good reflections of the real situation
treatment is also of important concern for upgrading. The oil phase of (Fig. B.7).
the reclaimed oil after steam injection contained almost the same ratios Because the density of S-OS was lower than 1 g/cm3, whereas that of
of SARA fractions and element compositions (C, H, N, and S) with the B-OS was higher than 1 g/cm3 (Table 1). The floating separation of
extracted oil by direct azeotropic distillation (Fig. 6). Since the tem- layered oily sludge in water was supposed to be spontaneous, but water
perature of the injection area directly contacting the oil layer remained was hard to penetrate into the layered oily sludge at ambient tem-
at around 100 °C, the volatile matters of this reclaimed oil, mainly perature due to high viscosity of the oily sludge, making it a very long
contained in saturates, could be hardly released unlike steam distilla- process to float oil. When the steam jet was applied to the oil layer, it
tion [37,51]. As a result, this mild processing temperature has another took only 60 s to finish this floating process according to the simulation
advantageous feature to prevent light fractions of oil from volatiliza- results, where the time of duration was also within the range from the
tion. experimental results (Fig. 2b). It can be concluded that the strong
heating and shocking effect of the steam jet caused surface disturbance

Fig. 2. Effect of steam temperature and steam to oily sludge ratio (S/OS) on oil reclamation rate (ORR) (a) and required time (b).

5
D. Xia, et al. Fuel xxx (xxxx) xxxx

Fig. 4. Changes of steam dynamic pressure with operating temperatures (corresponding steam temperatures at the beginning of the stable injection stage are shown
in brackets).

Fig. 6. SARA analysis and element compositions of extracted oil before and
after steam jet process.

rushed into the intermediate water phase, contacting surrounding water

and re-floating to merge with the top oil layer. This purification process
continued until the steam jet was over.
For water-solvable metals, including Mg, K, Ca, Al, and Na, they are
usually present in combination with naphthenic acid [24]. The tiny oil
Fig. 5. Demetallization effect of the steam jet process. droplets provide sufficient surface area to facilitate the transfer of these
metal salts to the aqueous phase [26], which explained the preferable
removal rate of the light metals. For metallic impurities attached to
and significantly improved the oil’s spontaneous floating separation, solid particles, steam heating improves oil desorption from particles
leading to the favorable oil reclamation performance. owing to its positive enthalpy change as well as gravity settling of
Second, upon a floated oil layer was formed, oil purification took particles due to a reduction of oil viscosity [13,54]. Water-insoluble
effect. It can be seen from the numerical simulation of the oil pur- zinc compounds could be involved in this case because some Zn con-
ification process (Fig. 8), the intense steam jet had a strong impact on tents still remained in the extracted oil phase. As to Ni, V, and part of
the floated oil layer that was rapidly disintegrated into substantial tiny Fe, they are present as oil-soluble porphyrin-type compounds. Because
oil droplets. The number of oil droplets given by numerical simulation the oil fractions and C, H, N, S elements kept almost unchanged after
was even less than the real situation due to computational limits (Fig. the steam jet processing (Fig. 6), these metal porphyrins were scarcely
B.7). Then these droplets gaining kinetic energy from the steam jet possible to be removed by cracking with a formation of coke. Therefore,

6
D. Xia, et al. Fuel xxx (xxxx) xxxx

Fig. 7. Oil reclamation process simulated by COMSOL Multiphysics.

Fig. 8. Oil purification process simulated by COMSOL Multiphysics.

these metals could be involved in a reaction with steam to form water- water, resulting in concurrent effects of oil reclamation and purifica-
soluble compounds since the demetallization of metalloporphyrins by tion.
acid is a reversible action and a higher temperature increases ion-pro-
duct of water, and thus hydrogen ion [26]. Declaration of Competing Interests
From a practical point of view, the reclaimed oil in the upper layer
could be easily scraped out for upgrading while processing with the The authors declare that they have no known competing financial
steam jet. The condensed water in the middle layer should be re-con- interests or personal relationships that could have appeared to influ-
veyed to a steam generator after remediation. The residual oily sludge ence the work reported in this paper.
can be more suitable to be treated by traditional methods, such as
solvent extraction or thermochemical washing. Although this study Acknowledgments
shows that steam jet has great potential to treat layered oily sludge in
heavy oil production areas, its external configuration, such as the This research was financially supported by the National Nature
structure of nozzles and level of condensed water and oil layer, needs to Science Foundation of China (No. 51778606, No. 21477151) and the
be well designed to further optimize this process for a pilot trial [55]. National Water Pollution Control and Management Program (No.
2017ZX07107-002). The authors sincerely appreciate the collection
4. Conclusions support provided by Shengli Oilfield, China. We also would like to
thank Dr. Ying Zhou from the University of Chicago for providing many
Layering is a distinct phenomenon for oily sludge in storage ponds constructive comments.
of oil joint stations. Steam jet is efficient to reclaim and purify oil from
the layered oily sludge. Higher temperatures and S/OS generate better Appendix A. Supplementary data
oil reclamation performance and longer duration improves oil pur-
ification. However, taking the economic costs as a factor, the proper Supplementary data to this article can be found online at https://
condition was determined at 300 °C and S/OS of 6:1, where over 90% of doi.org/10.1016/j.fuel.2019.116731.
oil was reclaimed within 5 min and over 80% of metallic impurities
were removed while light fractions of oil phase remained. The nu- References
merical simulation reveals that steam jet with high temperature and
high dynamic pressure can expedite oil flotation and scatter oil layer, [1] da Silva LJ, Alves FC, de França FP. A review of the technological solutions for the
thus facilitate diffusion of metallic impurities of oil into surrounding treatment of oily sludges from petroleum refineries. Waste Manage Res

7
D. Xia, et al. Fuel xxx (xxxx) xxxx

2012;30:1016–30. https://doi.org/10.1177/0734242X12448517. [29] Khan MK, Kwek W, Kim J. Conversion of petroleum emulsion into light fraction-rich
[2] Hu GJ, Li JB, Zeng GM. Recent development in the treatment of oily sludge from upgraded oil in supercritical methanol. Fuel 2018;218:78–88. https://doi.org/10.
petroleum industry: a review. J Hazard Mater 2013;261:470–90. https://doi.org/ 1016/j.fuel.2017.12.122.
10.1016/j.jhazmat.2013.07.069. [30] Hart A, Greaves M, Wood J. A comparative study of fixed-bed and dispersed cata-
[3] Wang X, Wang Q, Wang S, Li F, Guo G. Effect of biostimulation on community level lytic upgrading of heavy crude oil using-CAPRI. Chem Eng J 2015;282:213–23.
physiological profiles of microorganisms in field-scale biopiles composed of aged oil https://doi.org/10.1016/j.cej.2015.01.101.
sludge. Bioresour Technol 2012;111:308–15. https://doi.org/10.1016/j.biortech. [31] Altunina L, Kuvshinov V. Improved oil recovery of high-viscosity oil pools with
2012.01.158. physicochemical methods and thermal-steam treatments. Oil Gas Sci Technol
[4] Huc AY. Heavy crude oils: from geology to upgrading: an overview. Paris (France): 2008;63:37–48. https://doi.org/10.2516/ogst:2007075.
Editions Technip; 2011. [32] Giacchetta G, Leporini M, Marchetti B. Economic and environmental analysis of a
[5] Im SI, Shin S, Park JW, Yoon HJ, Go KS, Nho NS, et al. Selective separation of Steam Assisted Gravity Drainage (SAGD) facility for oil recovery from Canadian oil
solvent from deasphalted oil using CO2 for heavy oil upgrading process based on sands. Appl Energy 2015;142:1–9. https://doi.org/10.1016/j.apenergy.2014.12.
solvent deasphalting. Chem Eng J 2018;331:389–94. https://doi.org/10.1016/j.cej. 057.
2017.08.094. [33] George CE, Azwell DE, Adams PA, Rao GV, Averett DE. Evaluation of steam as a
[6] He L, Lin F, Li XG, Sui H, Xu ZH. Interfacial sciences in unconventional petroleum sweep gas in low temperature thermal desorption processes used for contaminated
production: from fundamentals to applications. Chem Soc Rev 2015;44:5446–94. soil clean up. Waste Manage 1995;15:407–16. https://doi.org/10.1016/S0956-
https://doi.org/10.1039/c5cs00102a. 053X(99), 80029-4.
[7] Shin S, Lee JM, Hwang JW, Jung HW, Nho NS, Lee KB. Physical and rheological [34] Guo K, Li H, Yu Z. In-situ heavy and extra-heavy oil recovery: a review. Fuel
properties of deasphalted oil produced from solvent deasphalting. Chem Eng J 2016;185:886–902. https://doi.org/10.1016/j.fuel.2016.08.047.
2014;257:242–7. https://doi.org/10.1016/j.cej.2014.07.037. [35] Mishra SK. Separation process for treatment of oily sludge. US Patent 4851123;
[8] Bhattacharyya JK, Shekdar AV. Treatment and disposal of refinery sludges: Indian 1989.
scenario. Waste Manage Res 2003;21:249–61. https://doi.org/10.1177/ [36] Wang J, Han X, Huang Q, Ma Z, Chi Y, Yan J. Characterization and migration of oil
0734242X0302100309. and solids in oily sludge during centrifugation. Environ Technol 2017:1–9. https://
[9] Liang J, Zhao L, Du N, Li H, Hou W. Solid effect in solvent extraction treatment of doi.org/10.1080/09593330.2017.1329352.
pre-treated oily sludge. Sep Purif Technol 2014;130:28–33. https://doi.org/10. [37] McCoy DE. Nederland, Tex, Recovery of oil from refinery sludges by steam dis-
1016/j.seppur.2014.03.027. tillation. US Patent 4014780; 1977.
[10] Zubaidy EA, Abouelnasr DM. Fuel recovery from waste oily sludge using solvent [38] Cheng S, Wang Y, Fumitake T, Kouji T, Li A, Kunio Y. Effect of steam and oil sludge
extraction. Process Saf Environ Prot 2010;88:318–26. https://doi.org/10.1016/j. ash additive on the products of oil sludge pyrolysis. Appl Energy 2017;185:146–57.
psep.2010.04.001. https://doi.org/10.1016/j.apenergy.2016.10.055.
[11] Hu G, Li J, Hou H. A combination of solvent extraction and freeze thaw for oil [39] Huang Q, Mao F, Han X, Yan J, Chi Y. Characterization of emulsified water in
recovery from petroleum refinery wastewater treatment pond sludge. J Hazard petroleum sludge. Fuel 2014;118:214–9. https://doi.org/10.1016/j.fuel.2013.10.
Mater 2015;283:832–40. https://doi.org/10.1016/j.jhazmat.2014.10.028. 053.
[12] Jing G, Chen T, Luan M. Studying oily sludge treatment by thermo chemistry. [40] Wegner J, Ganzer L. Numerical simulation of oil recovery by polymer injection
Arabian J Chem 2016;9:S457–60. https://doi.org/10.1016/j.arabjc.2011.06.007. using COMSOL. In: Excerpt from the Proceedings of the 2012 COMSOL Conference
[13] Huang Q, Han X, Mao F, Chi Y, Yan J. A model for predicting solid particle behavior in Milan, Milan; 2012.
in petroleum sludge during centrifugation. Fuel 2014;117:95–102. https://doi.org/ [41] Sharma P, Kumar SS, Nashine B, Veerasamy R, Krishnakumar B, Kalyanasundaram
10.1016/j.fuel.2013.09.002. P, et al. Development, computer simulation and performance testing in sodium of
[14] Jean D, Lee D. Expression deliquoring of oily sludge from a petroleum refinery an eddy current flowmeter. Ann Nucl Energy 2010;37:332–8. https://doi.org/10.
plant. Waste Manage 1999;19:349–54. https://doi.org/10.1016/S0956-053X(99) 1016/j.anucene.2009.12.009.
00139-7. [42] Wilcox DC. Formulation of the k–w Turbulence Model Revisited. AIAA J
[15] Chen G, He G. Separation of water and oil from water-in-oil emulsion by freeze/ 2008;46:2823–38. https://doi.org/10.2514/1.36541.
thaw method. Sep Purif Technol 2003;31:83–9. https://doi.org/10.1016/S1383- [43] Batchelor GK. Bernoulli's theorem for steady flow of a frictionless non-conducting
5866(02)00156-9. fluid. In: Batchelor GK, editor. An introduction to fluid dynamics. New York:
[16] Zhang J, Li J, Thring RW, Hu X, Song X. Oil recovery from refinery oily sludge via Cambridge University Press; 2000. p. 156–64.
ultrasound and freeze/thaw. J Hazard Mater 2012;203:195–203. https://doi.org/ [44] COMSOL. Rising bubble. In: COMSOL Multiphysics 5.3.
10.1016/j.jhazmat.2011.12.016. [45] Smith DF, Rodgers RP, Rahimi P, Teclemariam A, Marshall AG. Effect of thermal
[17] Xu N, Wang W, Han P, Lu X. Effects of ultrasound on oily sludge deoiling. J Hazard treatment on acidic organic species from athabasca bitumen heavy vacuum gas oil,
Mater 2009;171:914–7. https://doi.org/10.1016/j.jhazmat.2009.06.091. analyzed by negative-ion electrospray Fourier transform ion cyclotron resonance
[18] Ramaswamy B, Kar D, De S. A study on recovery of oil from sludge containing oil (FT-ICR) mass spectrometry. Energy Fuels 2009;23:314–9. https://doi.org/10.
using froth flotation. J Environ Manage 2007;85:150–4. https://doi.org/10.1016/j. 1021/ef8005564.
jenvman.2006.08.009. [46] Kumar K, Nikolov A, Wasan D. Mechanisms of stabilization of water-in-crude oil
[19] Shie J-L, Lin J-P, Chang C-Y, Lee D-J, Wu C-H. Pyrolysis of oil sludge with additives emulsions. Ind Eng Chem Res 2001;40:3009–14. https://doi.org/10.1021/
of sodium and potassium compounds. Resour Conserv Recyl 2003;39:51–64. ie000663p.
https://doi.org/10.1016/S0921-3449(02)00120-9. [47] Mullins OC, Sheu EY, Hammami A, Marshall AG. Asphaltenes, heavy oils, and
[20] Liu J, Jiang X, Zhou L, Han X, Cui Z. Pyrolysis treatment of oil sludge and model- petroleomics. New York: Springer; 2007.
free kinetics analysis. J Hazard Mater 2009;161:1208–15. https://doi.org/10.1016/ [48] Long XW, Zhang GL, Han L, Meng Q. Dewatering of floated oily sludge by treatment
j.jhazmat.2008.04.072. with rhamnolipid. Water Res 2013;47:4303–11. https://doi.org/10.1016/j.watres.
[21] Gong Z, Du A, Wang Z, Bai Z, Wang Z. Analysis on integrated thermal treatment of 2013.04.058.
oil sludge by Aspen Plus. Waste Manage 2019;87:512–24. https://doi.org/10.1016/ [49] Zolfaghari R, Fakhru’l-Razi A, Abdullah LC, Elnashaie SS, Pendashteh A.
j.wasman.2019.02.038. Demulsification techniques of water-in-oil and oil-in-water emulsions in petroleum
[22] Heidarzadeh N, Gitipour S, Abdoli MA. Characterization of oily sludge from a industry. Sep Purif Technol 2016;170:377–407. https://doi.org/10.1016/j.seppur.
Tehran oil refinery. Waste Manage Res 2010;28:921–7. https://doi.org/10.1177/ 2016.06.026.
0734242X09345794. [50] Taleghani ST, Jahromi AF, Elektorowicz M. Electro-demulsification of water-in-oil
[23] Viana FF, de Castro Dantas TN, Rossi CGFT, Neto AAD, Silva MS. Aged oil sludge suspensions enhanced with implementing various additives. Chemosphere
solubilization using new microemulsion systems: design of experiments. J Mol Liq 2019;233:157–63. https://doi.org/10.1016/j.chemosphere.2019.05.161.
2015;210:44–50. https://doi.org/10.1016/j.molliq.2015.02.042. [51] Zhou L, Jiang X, Liu J. Characteristics of oily sludge combustion in circulating
[24] Ali MF, Abbas S. A review of methods for the demetallization of residual fuel oils. fluidized beds. J Hazard Mater 2009;170:175–9. https://doi.org/10.1016/j.
Fuel Process Technol 2006;87:573–84. https://doi.org/10.1016/j.fuproc.2006.03. jhazmat.2009.04.109.
001. [52] Kolev V, Kochijashky I, Danov K, Kralchevsky P, Broze G, Mehreteab A.
[25] Hart A, Leeke G, Greaves M, Wood J. Down-hole heavy crude oil upgrading by Spontaneous detachment of oil drops from solid substrates: governing factors. J
CAPRI: effect of hydrogen and methane gases upon upgrading and coke formation. Colloid Interface Sci 2003;257:357–63. https://doi.org/10.1016/S0021-9797(02),
Fuel 2014;119:226–35. https://doi.org/10.1016/j.fuel.2013.11.048. 00052-8.
[26] Hu GJ, Li JB, Thring RW, Arocena J. Ultrasonic oil recovery and salt removal from [53] Dai Q, Chung KH. Hot water extraction process mechanism using model oil sands.
refinery tank bottom sludge. J Environ Sci Health A 2014;49:1425–35. https://doi. Fuel 1996;75:220–6. https://doi.org/10.1016/0016-2361(95), 00218-9.
org/10.1080/10934529.2014.928556. [54] Li Z, Wu P, Hou X, Liu D, Wang J, Lou B, et al. Probing the essence of strong
[27] Abbas S, Maqsood Z, Ali M. The demetallization of residual fuel oil and petroleum interaction in oily sludge with thermodynamic analysis. Sep Purif Technol 2017.
residue. Pet Sci Technol 2010;28:1770–7. https://doi.org/10.1080/ https://doi.org/10.1016/j.seppur.2017.06.044.
10916460903226122. [55] Chong D, Zhao Q, Yuan F, Wang W, Chen W, Yan J. Research on the steam jet length
[28] Khan MK, Sarkar B, Zeb H, Yi M, Kim J. Simultaneous breaking and conversion of with different nozzle structures. Exp Therm Fluid Sci 2015;64:134–41. https://doi.
petroleum emulsions into synthetic crude oil with low impurities. Fuel org/10.1016/j.expthermflusci.2015.02.015.
2017;199:135–44. https://doi.org/10.1016/j.fuel.2017.02.070.