Article pubs.acs.org/IECR
Experimental Study of Silver Cathode for Electrochemical Deoxygenation of Seawater for Enhanced Oil Recovery Utsav R. Dotel,†,‡ Kai Vuorilehto,§ Magne O. Sydnes,† Hans Urkedal,‡ and Tor Hemmingsen*,† †
Department of Natural Science and Mathematics, University of Stavanger, NO-4036 Stavanger, Norway Deox AS, Professor Olav Hanssens vei 7A, NO-4021 Stavanger, Norway § Department of Chemistry, Aalto University, 02150 Espoo, Finland ‡
ABSTRACT: Seawater injection into reservoirs is a common method to enhance oil recovery. In order to avoid corrosion in pipes, the oxygen has to be removed from the seawater. This study focuses on the use of silver cathodes in electrochemical cells for oxygen removal, and the physical parameters that affect the efficiency of this process. Three electrochemical cells have been constructed and tested with packed bed cathodes made up of pure silver particles and silver plated brass spheres. Experimental results showed that silver was a suitable cathode material for oxygen removal. The most important parameter for optimizing cell performance was the depth of the cathode chamber. In addition, the thickness of the silver plating is also an important parameter to adjust since a too-thin layer results in erosion giving rise to galvanic corrosion between brass and silver, which reduces the efficiency of the cell.
1. INTRODUCTION Corrosion is an important and costly problem for pipelines and process systems on offshore platforms used for oil and gas production.1,2 There are numerous examples of severe corrosion attack in pipelines used for seawater injection.3,4 Dissolved oxygen in seawater plays a major role in the corrosion process. Thus, removing oxygen prior to injection would significantly reduce the corrosion.1,5 There are some methods available for oxygen removal from seawater offshore.6 The most commonly used chemical methods include the use of oxygen scavengers such as sulfite, metabisulfite, and bisulfite. Oxygen removal to a concentration less than 10 ppb can be achieved with the use of oxygen scavengers.7 However, many of the chemicals for oxygen removal are not environmentally friendly, and some chemicals can have a severe negative impact on the occupational safety of the personnel involved in the process, and may in addition increase corrosion when used in too high doses.8 The scavenging ability of these chemicals depends on pH and temperature, and catalysts might be required in order to have a sufficient reaction rate. Physical methods include oxygen stripping in vacuum towers or use of nitrogen gas purging. However, these methods only reduce the oxygen level to around 50 ppb under optimal conditions, and further use of chemicals might be necessary in order to reduce the oxygen concentration to a desirable level. High capital cost is another important issue with these techniques.9 Electrochemical removal of oxygen from water has received some interest, although most research has been directed toward © XXXX American Chemical Society
oxygen reduction in electrochemical energy conversion in fuel cells.10−12 To our knowledge no electrochemical techniques for oxygen removal have been developed at the industrial scale in order to remove dissolved oxygen and by such means avoid corrosion in pipelines. The commercial development of an energy efficient electrochemical method for oxygen removal would be beneficial by reducing the capital cost in addition to being environmentally friendly. An improved electrochemical system for oxygen removal from seawater is presented herein. Parameters such as flow rate, chemistry, and geometry of the electrochemical cell play a significant role in oxygen removal. A low overpotential was needed for the oxygen reduction reaction (ORR) using a threedimensional packed silver bed electrode. Materials such as copper, nickel, and graphite have been tested in previous studies, but shortcomings such as the corrosive property in seawater for copper and nickel, and the formation of hydrogen peroxide for graphite, make them less attractive as cathode materials.10 Platinum, gold, and palladium are too costly for large scale industrial deoxygenation, while silver with kinetics and oxygen reduction mechanism similar to platinum will be a better material candidate for cost-effective oxygen removal.13−15 Silver was therefore chosen as cathode material in the cells studied. Received: April 19, 2016 Revised: July 6, 2016 Accepted: July 20, 2016
A
DOI: 10.1021/acs.iecr.6b01493 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research
2. MATERIALS AND METHODS 2.1. Electrochemical Cells. Three types of electrochemical cells (cell A, cell B, and cell C) were designed with different dimensions. A schematic drawing of the cells is depicted in Figure 1. The cells were rectangular flat structures made of
Table 2. Compositions (in %) of Titanium Grade 1 and Grade 7 Alloys
acrylic glass. The purpose of having three cells with different dimensions was to compare the oxygen removal efficiency and observe the impact of cell size on oxygen removal properties. The dimensions of cell A, cell B, and cell C are presented in Figure 1 and Table 1. Table 1. Dimensions of Electrochemical Cells and Cathode Chambers (in cm) width (W)
depth (D)
length (L)
cathode width (Wc)
cathode depth (Dc)
cathode length (Lc)
A B C
31.0 31.0 18.5
2.9 3.0 4.5
31.0 51.0 32.0
24.5 24.3 12.5
1.15 0.50 0.40
20.0 39.0 24.0
titanium grade 1
C Fe H N O Pd Ti
max 0.1 max 0.2 max 0.015 max 0.03 max 0.18 0.0 99.5
titanium grade 7 max max max max max 0.2 99
0.1 0.3 0.015 0.03 0.25
weight) giving a total surface area of approximately 27 200 cm2. The cathodic volume was 563 cm3. The void volume was measured to be 33% using water displacement method. Electrochemical cell B was also a two-chamber electrochemical cell with anode and cathode material similar to cell A. Around 729 000 spheres were used in cell B giving a total surface area of about 22 900 cm2. The cathodic volume of cell B was 473 cm3 with a void volume of 33%. Electrochemical cell C was a three-chamber electrochemical cell, with one cathode chamber in between the two anode chambers. Cell C had the smallest depth of the cathode chamber among the three cells used in this work. Stainless steel AISI 316 was used as current feeder to the cathode, and 0.3− 0.6 mm irregular silver particles were used as cathode material in cell C. The total volume of silver particles was 130 cm3 including interparticle volume. Cell C was used for studying the formation of side products in the deoxygenation process. An analytical balance (VWR LAG 414i) was used for measuring the mass of the silver plated brass spheres before and after the experiments. A scanning electron microscope (Zeiss SUPRA 35 VP) with EDX and an optical microscope (Leica MZ 16) were used for analyzing the surface of the electrode material. A cation exchange membrane (Lewatit MC-3470) was used to allow cations to pass from the anode chamber to the cathode chamber. The thickness of the membrane was 0.051 cm, and the water permeability through the membrane was 269 mL h−1 m−2 at 0.3 bar according to the supplier.19 The membrane was immersed in 0.5 M sodium chloride solution for at least 2 h prior to insertion into the cell. For each experiment the system was flowed for 30 min in order to have stable conditions prior to the oxygen removal experiments. 2.2. Experimental Setup and Procedure. The test setup was built in the laboratory with a pump, flow meter, oxygen sensor, and pressure gauge at the inlet and outlet (Figure 2). Laboratory power supply (Manson HCS 3300) was used to maintain the required voltage between the anode and cathode. WTW Oxi 3315 was used as an oxygen sensor, and WTW Multi 3401 was used for conductivity measurements. All instruments were calibrated prior to operation following the recommended calibration procedure given by the supplier. The detection limit for the oxygen sensor was 1 ppb and the error limit was ±0.5% of the displayed value.20 Both a 0.5 M NaCl solution and raw seawater, taken from 80 m depth (unfiltered) at the research facilities of International Research Institute of Stavanger (IRIS) in Mekjarvik, Norway, were used as catholyte in these experiments. The conductivity of the 0.5 M NaCl solution was 48 mS cm−1, and the conductivity of seawater was 52 mS cm−1. The experiments were carried out at 20 ± 3 °C. A 0.05 M sodium sulfate solution was prepared as anolyte for cell A and cell B, whereas 0.01 M nitric acid was used as anolyte for
Figure 1. (a) Electrochemical cell dimensions. (b) Side view of cell A and cell B. (c) Side view of cell C.
cell
component
Electrochemical cell A was a two-chamber electrochemical cell with an anode and a cathode chamber, separated by a cation exchange membrane (CEM), Lewatit MC-3470, manufactured by Lanxess. Titanium (grade 1 alloy) mesh coated with iridium− tantalum based coating was used as the anode. The coating was performed by Swedish electrode material supplier Permascand AB where the titanium surface was cleaned, grit blasted, and etched. Then the coating precursor was applied and heated in order to decompose the precursor, thus forming a noble metal oxide on the substrate. The process of coating and heat treatment was continued until a layer thickness of 5 μm was obtained.16 The frame and the current feeder for the anode were made from titanium (grade 7 alloy). The standard compositions of titanium grade 1 and grade 7 alloys are presented in Table 2.17 Iridium oxide was chosen as anode due to its high corrosion resistance and the fact that it is a proven anode material for seawater electrolysis.18 The size of the anode inclusive frame was equal to the size of the cell. The thickness of the anode was 0.1 cm, and the size of the opening in the mesh was 1.0 cm × 0.3 cm. Silver plated brass spheres were used as a threedimensional packed bed cathode. The brass spheres were electroplated with silver at the production facility of Velo Galvanisk Anlegg AS, a Norwegian company involved in surface treatment and electrolysis. The sphere diameter was 0.1 cm, and the thickness of the silver layer was 2 μm. The approximate number of spheres used in cell A was 867 000 (based on B
DOI: 10.1021/acs.iecr.6b01493 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research
Figure 2. Experimental setup for the deoxygenation cell. The same setup was used for all three cells.
assumed to be consumed by the hydrogen evolution reaction, shown in eq 4. The resulting hydrogen mass fraction, CH2, was calculated using eq 6, where Mhyd is the molecular weight of hydrogen (g mol−1), Imeas is the measured current (A), Zhyd is the number of electrons required per molecule, and m is the mass flow rate (g s−1).
cell C. The anolytes were stagnant. Nitric acid was used instead of sodium sulfate in the anodic chamber in cell C in order to keep the pH stable. The oxygen concentration was measured continuously at the outlet of the cell. The anolyte tank was used to supply the anolyte in the cell after the used anolyte had decreased the pH. The hydrogen peroxide content was measured colorimetrically using Chemetrics analysis ampules.
C H2 =
3. RESULTS AND DISCUSSION When oxygen is removed electrochemically from water, molecular oxygen is reduced at the cathode side of the cell while water is oxidized at the anode side. The chlorine is usually evolved rather than water being oxidized to oxygen when chloride ion is present in the electrolyte. The possible cathode reactions with standard reduction potentials (SHE) are given in eqs 1−4. Hydrogen ions migrate through a cation exchange membrane from the anode chamber to the cathode chamber and combine with oxygen molecules under reductive conditions to form water or hydrogen peroxide as shown in eqs 1 and 2. In acidic media, oxygen reduction is shown to follow a fourelectron or a two-electron pathway.21 The ORR mechanism for peroxide formation is discussed in more detail in several papers.22−26 O2 + 4H+ + 4e− ⇌ 2H 2O
E° = 1.229 V
(1)
O2 + 2H+ + 2e− ⇌ H 2O2
E° = 0.695 V
(2)
H 2O2 + 2H+ + 2e− ⇌ 2H 2O +
−
2H + 2e ⇌ H 2
E° = 1.776 V
E° = 0.000 V
P = IV
(6)
(7)
Three cells with different geometries were used in order to study the effects of different physical parameters. In cell A the oxygen removal was studied by pumping water with four different salinities (0.01, 0.05, 0.1, and 0.5 M NaCl) at constant flow rates (500 and 750 mL min−1). Similarly, the effect of different flow rates on resulting cathodic currents and on oxygen removal efficiency was measured at constant applied voltages (2.2 and 2.3 V). Cell B was used for high efficiency deoxygenation tests. Physical parameters such as oxygen concentration, pressure, voltage, current, and conductivity were monitored during these test in order to reach higher than 99.9% deoxygenation of water. In cell C the potential side reactions that can take place during the deoxygenation process were studied. Also here more than 99.9% deoxygenation was achieved. Voltages in the range 1.7−2.2 V were applied in experiments with three different flow rates (250, 500, and 750 mL min−1). The side reactions at the cathode were examined by measuring hydrogen and hydrogen peroxide at the outlet of the cell. 3.1. Electrochemical Cell A. The efficiency of oxygen removal from water when applying 2.2 V increased with higher salinities as shown in Table 3 and Figure 3. Using a flow rate of 750 mL min−1 for a 0.01 M NaCl solution, the initial oxygen concentration of 9.4 ppm was reduced by 30% to 6.5 ppm, while for a 0.5 M NaCl solution the 8.6 ppm initial oxygen concentration was reduced by 57% to 3.7 ppm. The same test was performed at a lower flow rate of 500 mL min−1 for 0.01 M NaCl. The initial oxygen concentration of 9.4 ppm was reduced by 37% to 5.9 ppm, while for a 0.5 M NaCl solution with 8.6 ppm initial oxygen concentration, the oxygen concentration was reduced by 76% to 2.1 ppm.
(3) (4)
CO2Z O2Fm MO2
Z hydFm
The power (P) needed was calculated using eq 7 for the electrochemical cell, where I is the current (A) and V is the voltage (V).
For each experiment, the theoretical current (Itheor) was calculated using eq 5, where CO2 is the mass fraction of dissolved oxygen in the water (g/g), ZO2 is the number of electrons required per oxygen molecule (4), F is the Faraday constant (96 500 A s mol−1), m is the mass flow rate (g s−1), and MO2 is the molecular weight of oxygen (g mol−1). Itheor =
Mhyd(Imeas − Itheor)
(5) 9
In a study by Tamminen et al., the voltage was adjusted between the electrodes in the cell to give a 10% higher current than the theoretical current for oxygen reduction, and a similar procedure was used in our experiments. Thus, the cell efficiency was considered to be 0.9. The additional needed current was C
DOI: 10.1021/acs.iecr.6b01493 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research Table 3. Oxygen Removal for Different Concentrations of NaCl Solutions in Cell A final oxygen concn (ppm) NaCl solution (M)
initial oxygen concn (ppm)
0.01 0.05 0.1 0.5
9.4 9.5a 9.3 8.6
750 mL min−1 500 mL min−1 6.5 6.0 5.1 3.7
5.9 5.1 4.0 2.1
a
Note that the equilibrium oxygen concentration is generally higher for dilute solutions, but was measured lower for some electrolytes used due to variation in stirring and stabilizing time of the electrolytes prior to the experiments. This was also the reason the oxygen concentration in electrolyte was more than the saturation oxygen concentration.
Figure 4. Effect of flow rate for a 0.5 M NaCl solution on the output current at constant voltage (2.2 and 2.3 V).
In any electrochemical system, either current or voltage can be controlled. The voltage was controlled in these experiments in order to avoid undesirable reactions such as hydrogen evolution (at higher voltage) and hydrogen peroxide (at lower voltage). A constant voltage in the range from 2.2 to 3.0 V was applied on a 0.5 M NaCl electrolyte flowed at a rate of 750 mL min−1. Figure 5 shows the relationship between applied voltage and current and oxygen removal. The increase of voltage from 2.2 to 3.0 V yielded from 50 to 89% deoxygenation. Significant removal of oxygen was not achieved in cell A. One of the major reasons for this is the large depth of the cathode chamber, which allows the oxygen molecules in the cathode section to pass through without reduction. A simple explanation is that the larger cathode depth increases the distance of ions carrying ionic current and it results in larger voltage (IR) drop. However, the phenomenon related to impedance of a packed bed electrode is more complicated and has been addressed in many research papers.29,30 The variation of different parameters yields different oxygen removal capacities of the cell. 3.2. Electrochemical Cell B. In cell B, which had less depth of the cathode chamber, it was possible to achieve the removal of oxygen (99.7 >99.9 >98.7
concentrations of hydrogen peroxide and hydrogen at the outlet are shown in Table 6. Table 6. Hydrogen and Hydrogen Peroxide Formation (μg L−1) at Various Flow Rates and Voltages
Figure 5. Effect of voltage on current and oxygen removal at constant flow rate (750 mL min−1) in a 0.5 M NaCl solution.
0.25 L min−1 voltage (V)
solved in cell C by increasing the voltage and analyzing the outlet water for the formation of side products. The power consumption for more than 99.9% oxygen removal from seawater is not high. The power requirement, according to voltage and current consumption, in cell B is 1.61 W based on eq 7 for removal of more than 99.9% oxygen at a flow rate of about 560 mL min −1 . The energy for deoxygenation is calculated to be 0.05 Wh L−1. Table 5 shows the best results obtained for cell B and includes the oxygen removal efficiency and energy consumption at different flow rates. 3.3. Electrochemical Cell C. In cell C, different voltages and flow rates for oxygen removal were used in order to study the formation of hydrogen peroxide (H2O2) and hydrogen (H2) gas. At low voltage (1.7 V) hydrogen peroxide was formed and quantified using Chemetrics analytical kits. At high voltage, hydrogen evolution resulted as a side reaction. The
1.7 1.8 1.9 2.0 2.1 2.2 a
H2 5 50 80 − − −
H2O2 312 0 0 − − −
0.50 L min−1
0.75 L min−1
O2
H2
H2O2
O2
H2
H2O2
O2