RPSEA Report number: 11122-31
Interim Report to
Research Partnership to Secure Energy for America
RPSEA Contract Number: 11122-31
Development of Plasma Technology For Water Management of Frac/Produced Water
January 18, 2014
Principal Investigator: Young I. Cho
Professor
Department of Mechanical Eng. and Mechanics
Drexel University
3141 Chestnut Street, Philadelphia, PA 19104
1
RPSEA Report number: 11122-31
Introduction
Objective
The objective of the present project is to develop an integrated plasma water treatment system for
improved management of flowback or produced water from hydraulic fracturing of shale gas or
oil. The new system addresses multiple needs in frack/produced water treatment with three key
processes integrated into a continuous system, which consists of a plasma-induced softening
system, a plasma-assisted self-cleaning filtration, and vapor-compression distillation unit.
Drexel’s integrated plasma system attempts to resolve multiple issues in the handling and disposal
of large volumes of flowback water during the drilling phase and produced water during the
production phase.
Description of the Project
The present plasma water treatment technology consists of the following three components:
•
Plasma-Induced Water Softening
•
Advanced Filtration
•
Vapor-Compression Distillation (VCD)
Key Deliverables Associated with the Project
•
A new water softening technology utilizing plasma arc discharge
•
A self-cleaning filter utilizing spark discharge
•
An integrated plasma unit with VCD process for produced water treatment
Work Completed at Drexel University as of January 18, 2014
Task 1.0 -- Project Management Plan - Completed
Task 2.0 -- Technology Status Assessment - Completed
2
RPSEA Report number: 11122-31
Task 3.0 -- Technology Transfer - Technology transfer activities have been reported with each
monthly report.
Experiments (Task 4)
Task 4.0 -- Plasma-Induced Softening Process
The specific goal of Task 4 is to demonstrate the validity of a plasma-induced softening process
with produced water in laboratory tests at Drexel. The present plasma-induced softening process
utilizes a gliding arc generated between two circular electrodes with the top electrode used as
cathode and the bottom as anode. A compressed air is used to continuously move arc
circumferentially in the space between the two electrodes, which enters the space tangentially and
exits through a hole at the bottom circular disk electrode. Thus, a plasma arc jet is formed in
produced water, making direct contact with produced water.
A description of the work conducted under Task 4.0 follows:
Subtask 4.1. -- Modeling of Ca2+ Precipitation Process
When water is directly exposed to gliding arc discharge (GAD), the following reactions occur
with the dissociation of water molecules [1, 2]:
e + H2 O
e
+ H2 O
M+ +
H2 O
H 2 O+ + H2 O
+ e-
!
H + OH
!
H- + OH
(1-2)
!
H 2 O+
+ M
(1-3)
!
H3O+ + OH
(1-4)
(1-1)
Then, H2O2 is formed from the recombination of hydroxyl radicals [2-4]:
OH + OH + M
!
H 2 O2 +
M
(1-5)
In the overall set of reactions, the concentration of H2O2 in water increases with plasma treatment
[5, 6]. A significant pH drop observed in the present study can be attributed to positive charges
3
RPSEA Report number: 11122-31
(M+) created in the plasma discharge that reach the water molecules and exchange charges with
the water molecules, resulting in the creation of H3O+ ions and OH radicals playing an important
role in oxidation and sterilization [3, 6-8] through Equations (1-4) to (1-5).
When air is the carrier gas, nitrogen oxides can be formed from gas phase reactions of dissociated
nitrogen and oxygen [9-11]
e ! 2N + e-
(1-6)
O2 + e ! 2O + e-
(1-7)
N + O ! NO
(1-8)
NO + O ! NO2
(1-9)
N2 +
The nitrogen oxide (NO2) affects the pH of the water through the formation of acids and ions as in
the following reactions. The reaction between NO2 formed as shown in Equation (1-8) and
hydroxyl radicals (OH) formed as shown in Equations (1-1) to (1-4) can generate HNO3 which
results in acidic water (i.e., HNO3). [10, 11]
NO2 +
OH ! HNO3
(1-10)
A number of studies on the antimicrobial effects of plasma have indicated OH and NO as the
effective species. [2, 12-15] The generation of these two reactive species depends on the
humidity of air injected into the GAD [16, 17]. For example, the concentration of OH generated
by the GAD can increase much higher than that of NO as the humidity of the carrier gas increases.
The OH formed from the dissociation of water can lead to the formation of H2O2 due to the
recombination of OH. Hence, when pure water is injected into the exit nozzle of the GAD, the
concentration of H2O2 becomes much higher than that of nitrous acid (HNO3) or nitric acid
formed from the reaction of OH and NOx. [11, 16]
The present plasma water treatment system produced an increase in H2O2 concentration and a
low-pH environment in water. These two effects described in this study are present as the main
mechanism for oxidation and bacterial inactivation with water and air injected into a GAD. Note,
4
RPSEA Report number: 11122-31
that H2O2 is not itself a strong oxidizer. Also, note that acidic water is not a strong oxidizer.
However H2O2 in the presence of acidic water becomes a very strong oxidizer and an effective
tool for the oxidation and inactivation of microorganisms in plasma water treatment. [14, 18]
Mechanism of Plasma-Induced Softening Process
This model is also able to explain the mechanism by which plasma discharges precipitate dissolved
calcium ions to suspended calcium particles. There are three reactions that control the rate at which
dissolved calcium and bicarbonate ions combine and crystallize [19, 20]. The arc discharge
produces intense and highly localized focal areas of heating where the arc of temperature 2,000-­‐
3,000 K makes contact with water. Stochastic heating generated by arc plasma is efficient and
highly localized, dissociating the bicarbonate ions producing 𝑂𝐻! ions (see Reaction 1), a process
which upregulates the precipitation reaction of calcium carbonate (see Reactions 2 and 3) [19].
HCO3− (aq ) ⇔ OH − (aq ) + CO2 ( g )
(Re action 1)
endothermic process
OH − (aq ) + HCO3− (aq ) ⇔ CO32− (aq ) + H 2O(l ) (Re action 2) exothermic process
Ca 2+ (aq ) + CO32− (aq ) ⇔ CaCO3 ( s)
(Re action 3) exothermic process
The thermochemistry of the above three reactions in this model, leading to the precipitation of
dissolved calcium ions to CaCO3, can be explained as follows: Due to the positive Gibbs free
energy ΔG of +43.6 kJ/mol (i.e., an endothermic process), Reaction 1 cannot take place
spontaneously [19]. However, a high temperature source can dissociate bicarbonate ions (HCO3), producing OH- and subsequently precipitating dissolved calcium ions to CaCO3 via Reactions 2
and 3, whose Gibbs free energies are both negative (i.e., spontaneous reactions).
Thus, by simply increasing water temperature (i.e., volume heating), one can precipitate the
calcium ions from hard water due to the inverse solubility of calcium ions in water [19, 21].
Technically sound and simple, but bulk heating is not actually a feasible solution for water
softening due to high relative costs associated with volume heating, which can be approximated
as 360 kJ/L [22]. Instead of volume heating, a model is developed as part of the present study to
dissociate bicarbonate ions by creating plasma arc discharge of 2000-3000 K [23] in produced
5
RPSEA Report number: 11122-31
water such that the arc makes direct contact with produced water, a process which we refer to as
stochastic heating.
Next, in this chemical physics model, we calculate the values for OH- from both the volume
heating and stochastic heating in order to show the incremental benefit of stochastic heating in
the precipitation of calcium ions. The amount of OH- which one can produce from Reaction 1
per unit time can be calculated as [24]:
nOH − = nHCO − × k
3
where nHCO − is the number of HCO3- participating in Reaction 1, and k is the reaction rate
3
coefficient. According to the Arrhenius equation [20], the reaction rate coefficient k becomes
k = Ae − Ea / T
where Ea is activation energy, T is the system temperature (in the unit of eV or Kelvin). Due to
the exponential curve of the equation, the Arrhenius equation indicates that the higher the water
temperature is, the faster the reaction will be. Hence, one can expect an intense highly localized
heating of a small volume of water in a field of continuous points around the arc of 2000-3000
K, significantly and efficiently raising the temperature of water near the arc.
A critical aspect of this model is to determine whether or not the arc discharge to be used in the
present study can dissociate HCO3- without spending a relatively large amount of electrical
energy. Below we examine two cases (i.e., volume heating and stochastic heating) to find out
which case produces more OH -­‐ for the same amount of energy spent.
Case 1. Volume Heating
We assume to heat the entire volume of water by one degree (e.g., from 300 K to 301 K). Then,
the number of OH- we can produce for Ea ≈1 eV (i.e., 11,000 K) becomes [20]:
6
RPSEA Report number: 11122-31
nOH − = nHCO − × k = nHCO − × Ae − Ea / T = AnHCO − e −11000 / 301 = e −36.5 AnHCO −
3
3
3
3
Case 2. Stochastic Heating Using Plasma Discharge
We assume to heat 1% of the entire water volume by 100 degrees (e.g., from 300 to 400 K)
based on the order of magnitude of the size of gliding plasma arc jet in a water volume generated
by the present plasmatron with compressed air. The number of HCO3- ions participating in the
'
reaction is 1%, i.e., nHCO
= 0.01× nHCO3 , because the plasma arc discharge is assumed to heat
3
only 1% of the total water volume. Then, the number of 𝑂𝐻! ions we produce for Ea ≈1 eV (i.e.,
11,000 K) becomes [20]:
− Ea / T
'
'
'
'
nOH
= 0.01AnHCO3 e −11000 / 400 = 0.01e −27.5 AnHCO3 = e −32 AnHCO3
− = n HCO × k = n HCO × Ae
3
3
'
'
Comparing the number of the hydroxyl ions produced for the two cases, i.e., nOH and nOH
− , we
−
can see simplistic illustration that stochastic heating by a plasma arc discharge would produce
about 100 times more OH-, thus, dissociating bicarbonate ions and precipitating dissolved
calcium ions 100 times more efficiently than bulk volume heating in hard produced water brines.
The present model relies in no small part upon a mechanism of plasma-induced water softening
wherein plasma discharge produces highly localized focal areas of sharply increased temperature
around the arc, dissociating HCO3- into OH -­‐ and CO2 as shown in the above Reaction 1,
efficiently without spending a large amount of energy. In addition, the arc also produces
hydroxyl ions. Once the hydroxyl ions are present in water, the above Reactions 2 and 3 occur
spontaneously, resulting in the removal of bicarbonate ions (i.e., see Reaction 2) and at the same
time, the precipitation of dissolved calcium ions in the form of suspended calcium salt particles.
These particles can be removed using a plasma-assisted self-cleaning filter such that water
hardness is reduced from flowback and produced waters without frequent filter replacement or
cleaning.
Note, that in the produced water samples obtained at Drexel, the concentration of bicarbonate
ions is often significantly less than that of dissolved calcium ions [25]. When bicarbonate ions
7
RPSEA Report number: 11122-31
from produced water are consumed through Reactions 1 and 2, the precipitation of calcium ions
ceases. Accordingly, based on a majority of the produced water specimens tested at Drexel, a
significant concentration of calcium ions remain in produced water after bicarbonate ion
concentrations are significantly reduced. Based on this sequence of process steps described here
in Task 4.0, it is necessary for the present chemical physics model to verify whether or not the
problem of calcium-fouling can be prevented in plasma-treated produced water with excessive
amounts of calcium ions. Therefore, as part of the present project, we have investigated whether
or not one can prevent mineral fouling in a heat exchanger with plasma-treated produced water
with an excessive amount of dissolved calcium ions but no bicarbonate ions.
Following a range of experimental tests performed as part of the present study to verify the
abovementioned process step within the present model, this interim report presents that as long
as plasma discharge removes bicarbonate ions from produced water, the calcium-fouling
problem cannot occur. This is a major breakthrough and improvement from the current practice
of water softening of using lime (Ca(OH)2) and soda ash (Na2CO3) in produced water
management [25]. This breakthrough may have positive implications in industries other than
energy field water treatment, which utilize water softening techniques.
Subtask 4.2. -- Parametric Study of Ca2+ Precipitation Process at Power Supply Side
There are a number of parameters that affect the plasma-induced softening process, including: (a)
the type of plasma discharge employed, e.g., gliding arc or pulsed spark discharges; (b) current
and voltage levels of the power supply used to generate and discharge plasma; (c) the method of
application and delivery of the plasma discharge to water, i.e., whether arc is “transferred” or
“not-transferred” to the water volume; and (d) the volumetric flow rate of air or gas added, if any,
to facilitate plasma ignition in produced water; among others. Considering these variables and
others, we have developed an advanced plasma discharge to maximize the benefit of plasma water
treatment not only for water softening performance but also for oxidation and disinfection. One
of the most important variables in the generation of plasma discharge in produced water is the
high electric conductivity of produced water. Due to the high electric conductivity, the
conventional practice of plasma discharges developed for most conventional process waters did
not work in produced water. We present in this report how we have taken advantage of the high
8
RPSEA Report number: 11122-31
electric conductivity of produced water to generate even stronger more efficient plasma
discharges in produced water.
Spark Discharge in Produced Water
We conducted a series of tests using spark plasma discharges to examine the efficacy of the spark
plasma discharges on the treatment of produced water. Fig. 4.2.1 shows this test apparatus, which
consisted of two tanks (supply and receiving), a pump, flow meter, radio-frequency (RF) electric
field treatment reactor, plasma spark reactor, and a microbubble generator.
Fig. 4.2.1. Photograph of the test setup for spark plasma discharge treatment of produced water in
a continuous flow system.
9
RPSEA Report number: 11122-31
Fig. 4.2.2. Photographs of spark discharges in produced water. Since spark duration is about 10
microseconds, multiple images are provided taken at different times.
Subtask 4.3. Parametric Study of Ca2+ Precipitation Process in Water Side
Based on Drexel discussions with a nascent consortium of industry stakeholders, in our
perspective, there are two different methods of produced water treatment depending on operating
requirements: one is to disinfect flowback and produced water into clean brines for recycling (i.e.,
re-injection to wells) in hydraulic fracturing for shale oil and gas and the other is to distillate
produced water for reuse or to wastewater discharge standards. The first method only requires
disinfection to avoid corrosion (in pipes downhole) that might be caused by acid-producing
bacteria and sulfate-reducing bacteria in produced water.
The second method is the distillation of produced water, where feed water to the distillation unit
should be calcium-free in order protect capital equipment and ensure operating feasibility. Hence,
the distillation process can utilize a water-softening process, which includes ion exchange for
small applications and addition of scale-inhibiting chemicals in larger applications. The latter
10
RPSEA Report number: 11122-31
process is conducted in two stages. In the first stage of chemical water softening, barium,
strontium, and radium ions, which are considered hazardous materials, are removed via
precipitation using sodium sulfate, Na2SO4, as the sulfate ions readily coagulate and precipitate
these ions into respective sulfate salts [26]. Note that once these three ions are precipitate into
respective sulfate salts, they are no longer considered as hazardous materials [26]. In the second
stage of water softening, calcium and magnesium ions are removed by lime (Ca(OH)2) and soda
ash (Na2CO3), producing a large amount of sludge to be disposed via costly landfill processes.
After these water-softening processes are completed, the feed waters to distillation units have
little to no dissolved inorganics except sodium and chloride ions so that the mineral fouling does
not take place in heat exchangers in the distillation unit.
One important challenge is that produced water often has excessive concentrations of calcium
ions, sometimes well over 20,000 ppm. The present study focused on the fact that in spite of the
excessively high calcium ion concentrations in produced water, bicarbonate ion concentrations we
observed in our produced water samples were only around 4,000 ppm or less. In order for
calcium ions to create CaCO3 mineral fouling problems in the heat exchanger in the distillation
unit, calcium ions and bicarbonate ions must both be present. Although plasma discharge cannot
directly remove dissolved calcium ions from the produced water, it can relatively easily remove
the bicarbonate ions by the high temperature focal areas of plasma arc (i.e., 2000 – 3000 K). Next
we present our study for a new CaCO3 fouling prevention method using plasma discharge.
New Fouling Prevention Method using a Plasma Gliding Arc for Produced Water Treatment
Produced water is conventionally treated using a range of physical, chemical, and biological
methods. Since there are multiple needs that should be addressed in produced water treatment, a
variety of different conventional methodologies have been used, as recently reviewed by
Ahmadun et al. [25], including the following: activated carbon, various forms of filtration (such
as sand filters, cartridge filters, multi-media filtration, membrane filtration), organic-clay
adsorbers, chemical oxidation, UV disinfection, chemical biocides, air strippers, chemical
precipitation, water-softening by the application of lime soda, clarifiers, settling ponds, ion
exchange, reverse osmosis, evaporation, steam stripping, and acidification. In nearly all of the
above cases, each modality of technology typically achieves a single treatment target. For
11
RPSEA Report number: 11122-31
example, UV disinfection can only destroy bacteria and is unable to address any of the other
objectives. Similarly, Hayes and Arthur [27] described a number of processes that are potentially
applicable to produced water treatment, indicating that each process can be applied to only a
limited number of basic functions.
In this Subtask 4.3, we developed and demonstrated a new method for preventing scaling in
produced water. These studies were conducted in two parts, each with their own objectives, as
shown in Fig. 4.3.1: (1) to test a plasma-assisted method for bicarbonate removal in produced
water, and (2) to verify bicarbonate removal as a mechanism for scale prevention. For the first
objective, a plasma gliding arc discharge (GAD) [28-31] was applied to produced water samples
from hydraulic fracturing to demonstrate the softening of water and removal of bicarbonate
content. To meet the second objective, high concentrations of calcium ions and bicarbonate ions
ranging from zero to 500 ppm were artificially added to distilled water or municipal water. By
applying heat over 30 h to evaporate water, the effect of bicarbonate removal was verified as a
mechanism for CaCO3 fouling prevention. These experimental studies were designed to
demonstrate a new method using plasma GAD for pretreatment of complex waste waters for the
purpose of decreasing bicarbonate concentration and mineral fouling mitigation even in the
presence of high calcium ion concentrations.
Water hardness consists of permanent (i.e., calcium) hardness and temporary (bicarbonate)
hardness [19]. The plasma used in this study is ionized gas having concentrated energy with
highly localized temperature increases near the arc of the GAD [28]. The geometry of the GAD
system and its ability to distribute plasma in water makes it well suited to alter water chemistry,
including the bicarbonate ion concentration, in water. Although the present study focuses on the
ability of the GAD to prevent mineral fouling in produced water, the GAD can also oxidize
hydrocarbons and inactivate microorganisms in produced water, which are topics that will be
studied and reported in the future. Recently we reported that the GAD produced H+ ions in water,
thereby reducing the pH of water [5, 32]. The H+ ions react with bicarbonate ions in produced
water, converting them to H2O and CO2 (gas) [19]. Detailed geometry and specifications of the
GAD used in the treatment of produced water have been reported elsewhere [5, 32] and are not
repeated here.
12
RPSEA Report number: 11122-31
In many industries, the removal of temporary hardness by the addition of lime (i.e., CaO or
Ca(OH)2) is a common practice [19, 21, 25], and lime-softening of water is a highly effective
chemical means for preventing fouling. The drawback of adding lime is that it increases the solid
content of CaCO3 sludge, in effect increasing the overall cost of disposal to landfills or landfarms,
which is an important economic and environmental driver in the treatment of produced water
[25]. Since the removal of temporary hardness using plasma discharges has not been previously
reported in scientific literature or in industry publications, it is necessary to demonstrate the
efficacy of calcium carbonate scaling prevention in plasma-treated produced water.
Fig. 4.3.1. Flow sheet with fouling prevention study design and methods.
13
RPSEA Report number: 11122-31
Fouling Prevention -- Experimental Methods
Plasma gliding arc discharge configured for reverse tornado flow was utilized to generate air
plasma, which was applied to produced water for treatment. The plasma GAD reactor consisted
of a 1-L reservoir and a GAD plasma generator connected to a power supply and a compressed air
line. Feed water entered the reactor using a peristaltic pump. For safety, a snorkel ventilation
unit was employed directly above the setup for safe extraction of air and vapor from the system to
outside of the laboratory. For the first part of this study, produced water samples from hydraulic
fracturing for shale gas were used (Springville, PA).
At the beginning of each test, 700 mL of produced water was isolated into a 1-L beaker. At
baseline, 50 mL of this sample was extracted with a syringe, and laboratory assessments including
alkalinity and pH measurements were performed. The remaining 650 mL volume was loaded into
the cylindrical plasma reactor, and plasma power was turned on. The power supply parameters
for generation of plasma were recorded, including voltage and current. After 10 min, plasma
power was turned off, and 50-mL of treated water was sampled for laboratory assessments
including bicarbonate ion concentration. These tests were repeated a total of six times, and the
results are reported in Table 4.3.1.
For six additional tests, plasma treatment was continued beyond 10 min until the bicarbonate
concentration was reduced to zero. Water samples of 50-mL volume were extracted every 10 min
for one hour. Laboratory assessments were performed on all extracted water samples to assess
the time required to achieve bicarbonate removal (i.e., time to zero bicarbonate) as reported in
Table 4.3.1. During testing water was re-circulated at a flow rate of 50- 200 mL/min using the
peristaltic pump. Bicarbonate concentration was determined based on the assumption that
alkalinity was due almost entirely to hydroxides, carbonates, or bicarbonates and furthermore that
over the measured pH range, alkalinity was specifically due completely to bicarbonate [19, 33].
While the first part of this study utilized produced water samples from a Marcellus shale gas well
as previously described, the second part of the study employed municipal or distilled water with
bicarbonate and calcium artificially added to tap or distilled water. This was done to verify that
bicarbonate removal might be able to serve as a mechanism for scale prevention. The use of
14
RPSEA Report number: 11122-31
produced water for the fouling component of this study, on the other hand, would have been
expected to result in oversaturation of NaCl, which could interfere with quantification of scale
formation.
A schematic diagram of the fouling test facility is given in Fig. 4.3.2, which consisted of a 1-kW
heating element to vaporize water, inside of a 3.8-L capacity boiler vessel made of stainless steel.
Makeup water was supplied from a 38-L water reservoir tank to the boiler vessel to replace the
vaporized water so that the water level in the vessel could be maintained constant during the
fouling test. Water vapor was disposed of using a laboratory snorkel ventilation unit. A floating
ball valve was utilized to control water flow from the reservoir tank to the vessel. A cutoff valve
between the reservoir tank and the ball valve was used such that water could be fed by gravity
through the floating ball valve during the test and closed at the end of the test for cleaning. The
floating ball valve controlled the flow of makeup water and maintained a constant water level in
the boiler vessel during the fouling test.
Snorkel Ventilation
Unit
Valve
(open)
38-Liter
Water Reservoir
Floating
Ball Valve
3.8 Liters
AC Supply
(120 V)
Vessel
(steel)
Heating Element
(1kW)
Variable
Transformer
Fig. 4.3.2. Schematic diagram of experimental setup for testing the scaling ability of water.
A cylindrical cartridge-type heating element was used having dimensions as follows: outside
diameter = 15.65 mm and length = 154 mm. The cylindrical cartridge-heating element was
installed in the boiler vessel through a port in the sidewall of the vessel. It should be noted that
the heating element had unheated sections at both ends (approximately 10 mm at the free end and
15
RPSEA Report number: 11122-31
25 mm at the fixed end; see sections marked X in Fig. 4.3.3). The heating element was connected
to a variable transformer (120 V AC) for switching on or off. The temperature of the water in the
vessel was maintained using the variable transformer to allow the 1-kW element to heat at its full
capacity.
Upon the completion of each fouling test, the variable transformer was used to turn the power to
the heating element off. Then, the cut-off valve between the tank and floating ball valve was
closed to prevent any additional water from flowing from the reservoir to the vessel. The tubing
between the floating ball valve and the reservoir tank was also disconnected for cleaning prior the
next test.
Fig. 4.3.3. Digital photographs of the heating element used in the present study before and after
fouling tests with a fixed calcium concentration of 5,000 ppm. Unheated sections of heating
element are marked as “X”.
16
RPSEA Report number: 11122-31
Prior to fouling tests, the following measurements were taken of the heating element: weight and
diameter at eight points along the axial direction each spaced 22 mm apart from 0 to 154 mm,
which were used as the baseline data. Upon the completion of the fouling test (at t = 30 h), the
power to the heating element was turned off, and water in the vessel was pumped out using a
peristaltic pump at a low flow rate, in order to minimize the physical disturbances upon any scale
that had formed on the heating element. The heating element was allowed to cool to room
temperature for 30 min before the element was removed. Then, the heating element was placed
on a white sheet of paper on the laboratory bench and allowed to dry for one hour before a
photograph was taken.
Subsequently, the change in weight from t = 0 to 30 h was determined in order to quantify the
magnitude of scale that had accumulated on the heating element. The amount of the accumulated
scale was measured using a laboratory balance (Model VB-302A, Virtual Measurements &
Control, Santa Rosa, CA). The thickness of the scale layer in the heating element was then
determined by measuring the diameter of the heating element before and after the fouling test at 8
equidistant points along the axial direction using a digital caliper (Cen-Tech model, Harbor
Freight Tools, Calabasas, CA).
After the completion of fouling tests and assessments of the heating element, small pieces of the
scale approximately 1 cm x 1 cm in size were removed from the element using a razor blade and
submerged in 200 mL of distilled water to ascertain whether or not the scale from the heating
element would dissolve in water over the duration of 24 h; upon seeing that the scale had not been
dissolved in the distilled water for any of the three test cases, it was concluded that the scale was
an insoluble compound such as CaCO3 and not a soluble salt such as sodium chloride (NaCl).
This is also strongly supported by the fact that the test fluid only contained two major chemicals,
CaCl2 and NaHCO3, with a limited amount of Na+ (see Table 4.3.2).
The desired composition of fouling test water used in the present study was achieved by mixing
an appropriate proportion of calcium chloride (CaCl2) and sodium bicarbonate (NaHCO3) as
17
RPSEA Report number: 11122-31
shown in Table 4.3.2. Water samples for fouling tests (a) and (b) used municipal water (Camden,
NJ), whereas that for fouling test (c) was prepared using distilled water.
To prepare the fouling test water samples, a large reservoir tank was filled with 38 L of water.
Anhydrous CaCl2 powder (96%, extra pure, Acros Organics) was added to the reservoir with
gentle stirring (done by hand using a stainless steel rod), and the CaCl2 in water solution was left
to dissolve for 10 min following artificial water-hardening procedures previously utilized by our
research group [34]. For fouling test (a), anhydrous NaHCO3 powder (USP/FCC, Fisher
Chemical) was slowly added (over the course of 10 min) into the reservoir, while gently stirring
with a long metal rod. This solution was then left to sit for 5 min.
Prior to initiating fouling test runs, water from the makeup reservoir filled the steel vessel to a
pre-determined line on its sidewall marking 3.8 L volume, at which point the floating ball valve
automatically restricted water flow from the reservoir to the vessel. The water level in the
makeup reservoir tank was recorded at the beginning of the test and recorded repeatedly at 6-h
intervals over the test duration of 30 h. At the end of each fouling test, the volume of the water
sample remaining in the reservoir was measured by emptying the remaining water into a
graduated cylinder.
Water for chemistry analysis was sampled at t = 0 h using a 50-mL syringe to extract a total 200mL sample from the steel boiler vessel at each sampling time point. The first sample was used for
initial laboratory assessments of water chemistry at baseline (i.e., t = 0 h) for alkalinity, calcium
hardness, total dissolved solids, salinity, conductivity, and pH. At t = 6, 12, 24, and 30 h, the
water level in the makeup reservoir was recorded, and then a 200-mL water sample was again
extracted from the vessel in a similar manner. Samples were taken from the middle of the vessel
volume while the heating element was turned on. Actual laboratory assessments of the water
samples were performed after water samples reached room temperature and always within 20 h of
the completion of each fouling test.
Water chemistry analyses of samples were conducted using a number of different measurement
equipment and techniques. Water hardness was measured using a HI 3842 Hardness Range (40018
RPSEA Report number: 11122-31
3000 mg/L) chemical test kit from Hanna Instruments (Woonsocket, RI). Alkalinity was
measured using a Carbonate Hardness/Alkalinity Test Kit from Salifert (Duiven, Netherlands).
The pH and electric conductivity were measured using a MP521 Type pH/ Electric Conductivity
meter provided by SANXIN (Shanghai, China). Total Dissolved Solids (TDS) was measured by
using a procedure similar to EPA 160.1 Total Dissolved Solids Test Method [35]. For the
evaporation component of the TDS measurement procedures, the weight of a porcelain
evaporating dish was measured before and after filling the dish with a 100-mL test water sample
which was then heated using a laboratory hot plate (model Cimarec 3, Thermolyne, Dubuque,
IA), to evaporate the liquid until solid particles were left inside the dish. The weight change of
the evaporating dish provided a measure of the amount of solids, and this quantity was divided by
the volume of water evaporated to give a value of TDS. Mass measurements were performed
using a precision laboratory balance (Sartorius GP603S, Data Weighing Systems, Elk Grove, IL).
Fouling Prevention -- Results
For the plasma component of this study the GAD was applied directly to produced water, and
observed was a decrease in the concentration of bicarbonate ions in produced water as shown in
Table 4.3.1. The baseline concentration of bicarbonate ions in produced water samples used in
the present study was in the range of 680-780 ppm. The post-treatment bicarbonate concentration
ranged from 0 to approximately 500 ppm during the present fouling tests. All 12 of the plasma
GAD tests reported in Table 1 showed a decrease in bicarbonate concentration for the produced
water samples following GAD treatment. This phenomenon was observed by measuring the
bicarbonate concentration at time, t = 0 min and t = 10 min. For six of these tests (tests 7-12),
after sampling at 10 min, plasma power was left on and allowed to continue running to investigate
whether or not complete removal of bicarbonate content was possible. Water samples were taken
for laboratory assessments every 10 min for a total of one hour of treatment. Plasma test 7
showed complete removal from 760 ppm to zero bicarbonate within 50 min; plasma tests 8, 9,10
and 11 showed complete removal from 747 ppm, 684 ppm, 752 ppm, and 685 ppm respectively
within 30 min; and, plasma test 12 showed complete bicarbonate removal from 685 ppm within
20 min. It can be observed from Table 1 that the treatment time required before complete
bicarbonate removal was seen, decreased by more than half from 50 minutes to 20 minutes from
test 7 to test 12. Reasons for this dramatic difference in the plasma treatment time required to
19
RPSEA Report number: 11122-31
completely remove bicarbonate content were not clear. Because of the use of compressed air to
drive the gliding arc, the GAD is a highly dynamic system. Additional research is necessary to
study the reason for these large differences in treatment time for achieving complete bicarbonate
removal. It is noteworthy that the plasma GAD was able to decisively reduce bicarbonate in water
and despite extremely high calcium ion concentrations in these produced water samples, when
this water interacts with the heated surfaces of heat transfer equipment, mineral fouling such as
that of CaCO3 should not occur. The secondary part of this study utilized fouling tests to clearly
show the benefits of supplying bicarbonate-depleted process water into a heat transfer system.
Table 4.3.2 shows the initial conditions for the three fouling test cases. The average baseline
calcium ion concentration for all three tests in the present study was 5,000 ppm, ranging from
4,800 to 5,200, whereas the bicarbonate concentrations varied from 7 to 445 ppm. For the water
prepared for fouling test (b), NaHCO3 powder was not added to municipal water, but the
concentration of bicarbonate ions was 118 ppm, reflecting the bicarbonate ions in the municipal
water. Although distilled water was used for fouling test (c), a small amount of bicarbonate ions
(i.e., 7 ppm) was already present in the distilled water. The pH values of the three water samples
varied slightly. Both electric conductivity and salinity were relatively high in all three water
samples due to the addition of CaCl2.
Test No.
Volume
(mL)
1
2
3
4 5 6 7
8
9
10
11
12
650
650
650
650 650 650 650
650
650
650
650
650
Plasma Treatment Time
0 (min)
10 (min)
Bicarbonate Concentration
(ppm)
699
467
685
475
778
560
745 370 685 458 685 203 760
364
747
347
684
317
752
472
685
292
685
83
Time to
zero
bicarbonate
(min)
---­‐-­‐
-­‐-­‐ -­‐-­‐ -­‐-­‐ 50
30
30
30
30
20
Voltage
(kV)
Current
(A)
2.1
2.1
2.1
2 2 1 2.1
2
2
2
2
2
0.1
0.1
0.1
0.1 0.1 0.2 0.1
0.1
0.1
0.1
0.1
0.1
Table 4.3.1. Plasma-induced bicarbonate removal in various produced water samples
20
RPSEA Report number: 11122-31
Initial Conditions
Test
CaCl2
(g)
(a)
(b)
(c)
530
530
530
Chemicals Added
NaHCO3
Water
(g)
(L)
53
0
0
38 [Tap Water]
38 [Tap Water]
38 [Distilled
Water]
Ca2+
(ppm)
HCO3(ppm)
4800
5200
5200
445
118
7
Measured Properties
TDS
pH
Conductivity
(ppm)
(mS/cm)
Salinity
(ppm)
18,600
15,800
12,000
15,000
15,000
15,000
6.73
7.22
6.55
23
22.8
21.9
Table 4.3.2. Initial conditions for each test including amount of chemicals added to achieve
desired compositions of samples, and measured properties of water after preparing samples
Figure 4.3.3 shows photographs of the heating element before and after the fouling test obtained
for all three test cases. The eight points along the axial direction of the heating element, at which
diameters were measured to quantify scale thickness, are shown in the photographs. Figure 4.3.4
shows variations in scale thicknesses in the three fouling test cases, measured at eight axial points.
In the case of fouling test (a), the scale thickness was approximately 1.85 ± 0.09 mm between
position #2 and #6, whereas in the case of fouling test (b), the scale thickness was approximately
0.17 ± 0.05 mm between position #2 and #6. In the case of fouling test (c), the scale thickness
approached zero.
Fig. 4.3.4. Comparison of scale thickness at 8 points (each 22 mm apart) along a 154-mm long
heating element for each of the three test cases.
21
RPSEA Report number: 11122-31
Table 4.3.5 provides the masses of accumulated scale and the average scale thicknesses over the
entire heating element for the three test cases. As expected, fouling test (a) had the largest scale
accumulated, whereas fouling test (c) approached zero accumulation of scale.
Test
(a)
(b)
(c)
Bicarbonate
Concentration
(ppm)
500
100
0
Accumulated Scale
(g)
9.97
0.97
-0.01
Average Thickness of
Scale
(mm)
1.85
0.17
0.02
Table 4.3.5. Results of the amount and average thickness of scale accumulated on heating
element in three different concentrations of bicarbonate ions: 0, 100, and 500 ppm.
Figures 4.3.5a-5d show variations in the total water volume, salinity, electric conductivity, and
TDS over time for the three fouling test cases. The total water volume linearly decreased with
time from the initial value of 38 L to approximately 12 L due to evaporation as the heating
element provided a constant thermal energy during the fouling test. On the other hand, the
salinity, electric conductivity, and TDS increased almost linearly with time as expected. Note that
the conductivity and TDS in all three test cases reached 110 mS/cm and 95,000 ppm, respectively,
at the end of the fouling tests, which are values that are often seen in produced water.
Figures 4.3.6a-6c show variations in pH, bicarbonate ion and calcium ion concentrations over
time for the three test cases. The pH dropped by 0.25 and 0.34 in fouling tests (a) and (b),
respectively, reflecting the loss of bicarbonate ions due to heating, whereas pH remained almost
constant in fouling test (c) because there were almost no bicarbonate ions to lose. The
bicarbonate ion concentration significantly dropped in the first 6 h in fouling tests (a) and (b) due
to the intense heating in water (i.e., see Fig. 4.3.6b). The calcium ion concentration almost
linearly increased with time for all three cases as expected.
22
RPSEA Report number: 11122-31
a) b) c) d) Fig. 4.3.5. a) Total Water Volume, b) Salinity, c) Conductivity, and d) TDS over time for three
different concentrations of bicarbonate ions: 0, 100, and 500 ppm.
a) 23
RPSEA Report number: 11122-31
b) c) Fig. 4.3.6. a) pH, b) Bicarbonate ion concentration, c) Calcium ion concentration over time for
three different concentrations of bicarbonate ions: 0, 100, and 500 ppm
Fouling Prevention - Discussion
As part of Subtask 4.3, the present fouling prevention study explored the potential for a robust
strategy utilizing high voltage plasma discharge to remove carbonate hardness (by reducing
bicarbonate concentration), while allowing for extremely high concentrations of calcium ions,
especially in produced water treatment and desalination. This approach to fouling prevention has
the potential application for use in the treatment of very complex wastewaters, which upon
undergoing a thermal process such as distillation, can result in profound fouling crystallization
problems with a negative impact on fluid processing equipment [36, 37]. The best plasma GAD
system in this study was plasma test 12 (see Table 1), which was capable of bringing bicarbonate
concentration of the 650 mL of water from almost 700 ppm to 83 ppm within 10 minutes. In
24
RPSEA Report number: 11122-31
addition, plasma test 12 showed that bicarbonate content could be completely removed within 20min GAD-treatment time.
The results presented in the fouling component of this study showed that given extremely high
concentration of calcium ions, i.e. 5,000 ppm, and bicarbonate level as low as 7 ppm, the
thickness of the CaCO3 layer on a 1 kw heater was under 0.03 mm as compared to a 1.85 mm
thick layer for the case of a 445 ppm bicarbonate concentration. The fouling component of this
study also provides experimental trendline data on the duration of time required to form scale
given the aforementioned initial conditions of sample waters and system operating conditions.
While it is well-known that CaCO3 scale is caused by the combination of calcium ions and
bicarbonate ions, to date, strategies for preventing formation of scale often emphasize removing
hardness by reducing calcium content such as through reverse osmosis, ion exchange, the addition
of soda ash (Na2CO3) or even electrochemical processes [21, 25, 38]. In produced water, the
concentration of calcium ions is often extremely high [25] such that its removal is generally
impractical. Accordingly, the removal of bicarbonate ions is an alternative strategy for the
purpose of softening. Conventionally, the removal of bicarbonate ions in hard water is
accomplished by adding lime (CaO) [19, 25]. However, this chemical approach increases the
overall mass of solids which ultimately need to be disposed reflecting a concomitant increase in
cost [25].
In this Subtask 4.3, an early feasibility study was performed for a new fouling prevention method
using a plasma discharge as a non-chemical device targeting bicarbonate in produced water.
Since the plasma arc discharge generates intense, highly localized heating [39-41], the arc can
been utilized to remove bicarbonate ions in produced water as shown in this study. The fouling
problems associated with CaCO3 particles are caused when both calcium ion and bicarbonate ion
are present in water [19, 21]. By dissociating and releasing bicarbonate, this plasma-based
methodology removes one of two key culprits in the fouling process without addition of other
solid materials as in chemical treatment. The present study utilized a non-chemical, clean
technology for removing bicarbonate ions in produced water. Subsequently, using fouling tests
25
RPSEA Report number: 11122-31
with artificially hardened municipal and distilled water, bicarbonate control was verified as a
mechanism for CaCO3 fouling prevention.
Our present mechanism for the modulation of bicarbonate by plasma in water is the effect of
intense, highly localized heating (or “stochastic heating”) which dissociates bicarbonate ions to
hydroxyl ions (OH-) and CO2 (gas) as described in our prior reports [23, 39]. Subsequently, the
hydroxyl ions react with other bicarbonate ions, producing carbonate ions (CO32-), which react
with Ca2+ to form CaCO3 particles [41, 42]. However in the present case, the generation of H+
ions in water by plasma GAD is also capable of removing bicarbonate ions by converting them to
H2O and CO2 (gas). This reaction in plasma treated produced water may provide the basis for a
new non-chemical fouling prevention method.
For the fouling tests, the bicarbonate ion concentrations significantly dropped in the first 6 h in
tests (a) and (b) (see Fig. 4.3.5b). Heating during these tests resulted in dissociation of
bicarbonate ions into OH- ions and CO2 gas, in a similar but less efficient manner than as
described in the present plasma mechanism. In the case of the fouling test, the last reaction of the
formation of CaCO3 particles takes place on the surface of the heating element in the form of
precipitation fouling [21]. Thus, although the bicarbonate ion concentrations were below 50 ppm
after t = 6 h, the CaCO3 particles suspended in water continued to adhere to the surface of the
heating element, increasing the scale thickness on the heating element over time in fouling tests
(a) and (b).
Summary of Task 4.3 Work
The present study investigated the feasibility of using plasma gliding arc discharge (GAD) to
prevent CaCO3 fouling in produced water. A two-part mechanistic approach was employed to
demonstrate the GAD plasma treatment method. The study demonstrated that the GAD could
effectively reduce the bicarbonate ions to zero in produced water. Subsequently, the effect of
bicarbonate modulation was verified as a mechanism for CaCO3 fouling prevention.
Additional research is necessary to investigate reasons behind the differences in required plasma
treatment time to achieve complete bicarbonate removal. Future studies should also include the
26
RPSEA Report number: 11122-31
following two topics: (1) a comprehensive experimental fouling study of scale thickness versus
time for given heat flux and initial water conditions; and (2) investigations of methods for
increasing the energy efficiency of plasma treatment such as plasma discharges with operating
parameters conducive to lowering bicarbonate concentration in produced water as a way of
potentially allowing for thermal treatment of sea water and wastewaters with a high mineral
content.
Development of Tube Blocking Test Apparatus
Dynamic tube blocking test, which is described in NACE standard method 31105, utilizes a
dynamic flow-through test apparatus and is used to determine the effect of bicarbonate
concentration on the formation of calcium carbonate scale with produced water in a heat
exchanger. Figs. 4.3.7 and 4.3.8 show the test system, which consisted of a coiled capillary tube
of inner diameter of 1.0 mm and length of 1.5 m made of 316 stainless steel and placed in a
heated oven maintained at 90°C during the tests. The tendency of tube blocking is measured by
measuring the pressure drop across the capillary tube over 8-15 h and recorded at PC. The flow
rate of produced water through the capillary tube is maintained by a peristaltic pump at 7 mL/min
according to the recommendation by the aforementioned NACE standard method. The test is
terminated when the increase in the pressure drop across the capillary tube exceeds a precalibrated level. After each test, the SS capillary tube is cleaned using 4 wt.% nitric acid until the
pre-calibrated pressure drop corresponding to a clean tube is restored.
27
RPSEA Report number: 11122-31
Fig. 4.3.7. Schematic diagram of tube blocking test apparatus: Pressure drop across coiled
capillary tube caused by scaling inside capillary tube is measured over time. NACE International
Publication 31105.
Fig. 4.3.8. Photograph of tube blocking test apparatus designed and constructed in the present
project: Pressure drop across capillary tube caused by scaling inside capillary tube is measured
over time. NACE International Publication 31105.
28
RPSEA Report number: 11122-31
Investigation of the Effectiveness of Radio-frequency (RF) Pulse Electric Fields on the Prevention
of Calcium Fouling in a Heat Exchanger using Tube-Blocking Test
The objective of the following test was to investigate whether radio-frequency (RF) pulse electric
fields can prevent calcium fouling in a heat exchanger. While this test was performed for the
purpose of fouling prevention in the produced water treatment system, it will be validated further
and integrated with the advanced filtration platform in a future Task. We constructed a oncethrough flow system shown in Fig. 4.3.9, where produced water with bicarbonate concentration of
approximately 820 ppm passed through a RF reactor at a flow rate of 200 mL/min. Note that the
RF reactor has two plate electrodes separated by 22 mm.
Fig. 4.3.9. Schematic Drawing of RF treatment. Water flow rate: 12 Liter/hour and
Fig.4.3.10 shows the test results of the aforementioned tube-blocking test. The capillary tube
blocked after 9.2 h in the case of untreated produced water, whereas it blocked after 21.5 h in the
case of untreated produced water. The concentration of bicarbonate ions in the untreated
produced water was 818 ppm, whereas that in the treated water was 419 ppm. Hence, the
significant delay in the calcium fouling in the capillary tube could be attributed to the reduction in
29
RPSEA Report number: 11122-31
the bicarbonate ions. Note that this significant delay took place in spite of excessively high
calcium concentrations of 3,000 ppm in the produced water.
Fig. 4.3.10. Tube-blocking time is lengthened in RF-Treated produced water from 9 to 22 h. The
concentrations of bicarbonate ions in produced water before and after RF treatment were 818 and
419 ppm, respectively. Note that Ca2+ concentration = 3,000 ppm.
Subtask 4.4 -- Oxidation of Hydrocarbons
When we apply arc discharge to produced water, the arc discharge oxidizes hydrocarbons, which
are in the form of suspended or dissolved organics such as BTEX (benzene, toluene, ethyl
benzene and xylene), PAH (polyaromatic hydrocarbons) and alkyphenols. Hydrocarbons in
produced water are able to make direct contact with high-temperature arc discharges in the
produced water medium. Accordingly, these hydrocarbons are effectively oxidized to H2O and
CO2 by the gliding arc plasma process. Plasma discharges at the attachment point to the water
surface have a “locally” very high temperature of 2000-3000 K that can provides ideal conditions
for vaporization and oxidation of hydrocarbons by atomic oxygen and active species of plasma
gas. Intensive mixing and transport of suspended and dissolved hydrocarbons to the plasma
region facilitate effective oxidation processes.
30
RPSEA Report number: 11122-31
Study of the Removal of Benzene from Produced Water
Tables 4.4.1 and 4.4.2 show BTEX composition in produced water (the highest values reported by
literature) and Analysis of Produced Water properties in referred Laboratory Experiments (water
from Wamsutter, Wyo) (Reference: [43]).
Benzene
Toluene
Ethylbenzene
Xylene
Formula
C6H6
C7H8
C8H10
C8H10
Boiling point
176.2°F(80.1°C)
231.1°F(110.6°C)
277 °F (136 °C)
281.3 °F(138.5 °C )
Density
876.50 kg/m³
866.90 kg/m³
866.50 kg/m³
864 kg/m³
Molar mass
78.11 g/mol
92.14 g/mol
92.14 g/mol
106.16 g/mol
Table 4.4.1. BTEX composition in produced water
mg/L
Volume to be
added
Ethyl
p-xylene
o-
benzene
/m-xylene
xylene
36.7
1.4
6.4
3.4
18 mL to
42 mL to
42 mL to
1.6 mL to
4 mL to
982
958
958
998.4mL(D.
996
mL(D.W)
mL(D.W)
mL(D.W)
W)
mL(D.W)
Benzene
Toluene
15.8
Table 4.4.2. BTEX composition in produced water prepared in the present test.
31
RPSEA Report number: 11122-31
Fig. 4.4.1. Test setup in the present project for the investigation of BTEX removal in produced
water (test procedure) using gliding arc discharge.
Fig. 4.4.2. Gas Chromatography analysis - BTEX degradation in produced water (test procedure)
32
RPSEA Report number: 11122-31
Fig. 4.4.3. Gas Chromatography analysis at DPI -- retention time: Benzene -17 min, Toluene -23
min, Ethylbenzene - 26 min.
Study of Disinfection of Produced Water with Plasma Discharges
Bacterial assessments were performed for acid-producing bacteria (APB) [44, 45] and sulfatereducing bacteria (SRB) [46, 47], which are commonly tested in oil and gas production
environments to monitor for risk of microbiologically influenced corrosion (MIC) [48]. The most
probable number (MPN) number was used to count active bacteria in water following NACE
standard TMO 194-2004: samples were incubated for 10 days at 35˚C in modified Postgate’s
medium B using for SRB testing and for 14 days at 35˚C in phenol red dextrose for APB testing
[49]. Modified Postgate’s medium B was used for SRB in order to complete tests in 10 days
instead of 30 days. For rapid assessments, adenosine triphosphate (ATP) was also measured.
Samples were passed through a 0.7-micron filter to collect microbiological cells and lysed.
Luciferase enzyme was added for bioluminescence, and a photometer was used to test
luminescence relative to a calibration fluid as a representative assessment of intracellular ATP.
33
RPSEA Report number: 11122-31
Stable discharge of gliding arc plasma was achieved. After treatment, samples are collected into
250 mL glass containers. Digital photographs of water samples were taken immediately after
tests, and changes in the color of originally black produced water were observed that were
dependent on treatment time. Gliding arc plasma treatment of produced water yielded a 4-log
reductions of both SRB and APB, demonstrating disinfection of bacterial sources of MIC from
water samples (see Table 4.4.3). In addition, 2-log reductions of total ATP levels were observed.
Figure 4.4.4 depicts ATP levels measured and their microbial equivalents as a function of gliding
arc discharge treatment time.
Figure 4.4.4. ATP Assessments of Produced Water with Gliding Arc Plasma Treatment.
Control
Bacteria
Assesments
ATP (Microbial
Equivalents/ mL)
SRB (CFU/mL)
APB (CFU/mL)
10
7.65
30 s
10
7.44
Plasma Treatment Time
1 min
10
7.48
5 min
105.07
104
104
104
< 100
104
104
104
100
Table 4.4.3 Bacteria Assessments of Produced Water with Gliding Arc Plasma Treatment. The
parameters of the plasma system include an air flow rate of 1 SCFM, and a power supply
operating at a current of 0.3 A, and voltage of 2 kV, for a power of 600 W.
34
RPSEA Report number: 11122-31
The gliding arc plasma discharge was demonstrated to disinfection of SRB and APB levels in
produced water in the laboratory setting and would be expected to reduce or inhibit MIC in in the
field. Energy requirement was 82 kJ/L per 1-log reduction.
Task 5 -- Plasma-Assisted Self-Cleaning Filtration.
The specific goal of this Task is to demonstrate the validity of plasma-assisted self-cleaning
filtration with produced water in laboratory tests. We have designed and constructed a tangential
flow filtration system using 0.2 micron membrane filter. The filtration study was conducted first
by measuring the variations of the head pressure of the membrane filter as a functin of flow rate.
Then, produced water was used to study whether the 0.2-micro filter can on its own remove any
compositions from the produced water.
Due to the delay in receiving the membrane filter from vendor, the self-cleaning function using
spark discharge could not be completed as of Dec. 31, 2013. The work is currently in progress.
Figure 4.5.1. Photograph and sketch of a tangential flow filtration system using 0.2 micron
membrane filter. Ceramic filter and Housing – Pore size: 0.2 micron
35
RPSEA Report number: 11122-31
Figure 4.5.2. Photograph and sketch of a filtration system using 0.2 micron membrane filter.
Measurement of pressure drop in ceramic filter. Water flows from inside to outside of filter.
400
375
Pressure Drop (mbar)
350
325
300
275
250
core to outside
outside to core
225
200
175
150
125
100
10
20
30
40
50
60
70
80
90 100 110 120 130 140 150
Water flow rate (mL/min)
Figure 4.5.3. Results from a filtration system using 0.2 micron membrane filter. Measurement of
pressure drop in ceramic filter.
36
RPSEA Report number: 11122-31
Figure 4.5.4. Sketch of a tangential flow filtration system using 0.2 micron membrane filter.
Measurement of pressure drop in ceramic filter at 0.8 LPM (300 gal/day). Water flows from
outside to inside of membrane filter.
6.4
Pressure drop (bar)
6.2
6.0
Pressure drop
5.8
5.6
5.4
5.2
5.0
0
20
40
60
80 100 120 140 160 180 200 220 240 260 280 300
Filtration Time (sec.)
Figure 4.5.5. Results from a filtration system using 0.2 micron membrane filter. Measurement of
pressure drop at 0.8 LPM (300 gal/day).
37
RPSEA Report number: 11122-31
Figure 4.5.6. Photograph of a tangential flow filtration system using 0.2 micron membrane filter
before and after filtration.
Figure 4.5.7. Photograph and sketch of a tangential flow filtration system using 0.2 micron
membrane filter combined with RF field treatment in the present study.
38
RPSEA Report number: 11122-31
Figure 4.5.8. Photograph of a tangential flow filtration system using 0.2 micron membrane filter
combined with RF field treatment in the present study.
Figure 4.5.9. Photograph and sketch of a membrane filter used in tangential flow filtration study
in the present study.
39
RPSEA Report number: 11122-31
Figure 4.5.10. Photograph of a small RF reactor used in tangential flow filtration study in the
present study.
Table 4.5.1. Results of water chemistry and biological measurements before and after filtration
using pulse radio-frequency (RF) electric field and plasma gliding arc discharge (600 W for 5 min
treatment) for produced water.
40
RPSEA Report number: 11122-31
Summary
For the past 8 months, we have pursued the best approaches for applying high-voltage plasma
discharges to produced water for optimum water management of flowback and produced water.
We have successfully created both gliding arc discharge and pulsed spark plasma discharges in
produced water. The maximum power of the present gliding arc discharge we have used thus far
is 600 W, whereas the maximum pulse energy of the present spark discharge is about 80 J per
pulse.
During the rest of the project period, we will design, construct, and test a vapor-compression
distillation unit. With the pretreatment of produced water using plasma discharges, we are
reasonably confident that we can operate the distillation unit without having calcium fouling
problem, the ultimate goal of the project.
References
1.
2.
3.
4.
5.
6.
7.
Oehmigen, K., et al., The role of acidification for antimicrobial activity of atmospheric
pressure plasma in liquids. Plasma Processes and Polymers, 2010. 7(3‐4): p. 250-257.
Fridman, A., Plasma chemistry. 2008: Cambridge Univ Press. Cambridge, U.K.
Locke, B.R. and K.Y. Shih, Review of the methods to form hydrogen peroxide in electrical
discharge plasma with liquid water. Plasma Sources Science and Technology, 2011. 20: p.
034006.
Medodovic, S. and B. Locke, Primary chemical reactions in pulsed electrical discharge
channels in water. Journal of Physics D: Applied Physics, 2009. 42: p. 049801.
Kim, H.-S., et al., Concentration of hydrogen peroxide generated by gliding arc discharge
and inactivation of E. coli in water. International Communications in Heat and Mass
Transfer, 2013. 42: p. 5-10.
Burlica, R., K.Y. Shih, and B. Locke, Formation of H2 and H2O2 in a water-spray
gliding arc nonthermal plasma reactor. Industrial & Engineering Chemistry Research,
2010. 49: p. 6342-6349.
Arjunan, K.P., et al., Non-thermal dielectric barrier discharge plasma induces
angiogenesis through reactive oxygen species. Journal of The Royal Society Interface,
2012. 9(66): p. 147-157.
41
RPSEA Report number: 11122-31
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
Joshi, A.A., et al., Formation of hydroxyl radicals, hydrogen peroxide and aqueous
electrons by pulsed streamer corona discharge in aqueous solution. Journal of Hazardous
Materials, 1995. 41(1): p. 3-30.
Sathiamoorthy, G., et al., KINETICS, CATALYSIS, AND REACTION ENGINEERINGChemical Reaction Kinetics and Reactor Modeling of NOx Removal in a Pulsed Streamer
Corona Discharge Reactor. Industrial and Engineering Chemistry Research, 1999. 38(5):
p. 1844-1855.
Burlica, R., M.J. Kirkpatrick, and B.R. Locke, Formation of reactive species in gliding
arc discharges with liquid water. Journal of Electrostatics, 2006. 64(1): p. 35-43.
Burlica, R., K.-Y. Shih, and B. Locke, Formation of H2 and H2O2 in a water-spray
gliding arc nonthermal plasma reactor. Industrial & Engineering Chemistry Research,
2010. 49(14): p. 6342-6349.
Gutsol, A., et al., Plasma for air and water sterilization. Plasma Assisted
Decontamination of Biological and Chemical Agents, 2008: p. 21-39.
Locke, B., et al., Electrohydraulic discharge and nonthermal plasma for water treatment.
Industrial & Engineering Chemistry Research, 2006. 45(3): p. 882-905.
Dobrynin, D., et al., Physical and biological mechanisms of direct plasma interaction with
living tissue. New Journal of Physics, 2009. 11: p. 115020.
Dobrynin, D., Physical and chemical mechanisms of direct and controllable plasma
interaction with living objects. 2011. Ph. D. Thesis, Drexel University, Philadelphia, PA.
Benstaali, B., et al., Density and rotational temperature measurements of the OH and NO
radicals produced by a gliding arc in humid air. Plasma chemistry and plasma processing,
2002. 22(4): p. 553-571.
Benstaali, B., et al., Plasma treatment of aqueous solutes: some chemical properties of a
gliding arc in humid air. EPJ APPLIED PHYSICS, 1998. 4: p. 171-180.
Yang, Y., et al., An underwater multi-channel plasma array for water sterilization.
Review of Scientific Instruments, 2011. 82: p. 096103.
Snoeyink, V.L. and D. Jenkins, Water chemistry. 1980, New York: John Wiley.
Weast, R.C., M.J. Astle, and W.H. Beyer, CRC handbook of chemistry and physics. Vol.
69. 1988: CRC press Boca Raton, FL.
Cowan, J.C. and D.J. Weintritt, Water-formed scale deposits. 1976: Gulf Publishing
Company, Book Division.
Bergman, T.L., et al., Fundamentals of heat and mass transfer. 2011: John Wiley & Sons.
Fridman, A. and L. Kennedy, Plasma Physics and Engineering. 2011, New York: Taylor
& Francis Group (2nd ed.).
Fridman, A. and L. Kennedy, Plasma Physics and Engineering. 2006, New York: Taylor
& Francis Group.
Ahmadun, F.l.-R., et al., Review of technologies for oil and gas produced water treatment.
Journal of Hazardous Materials, 2009. 170(2): p. 530-551.
Silva, J.M., Produced Water Pretreatment for Water Recovery and Salt Production
08122-36, RPSEA Final Report by General Electric Global Research 2012.
Hayes, T. and D. Arthur. Overview of emerging produced water treatment technologies. in
The 11th Annual International Petroleum Environmental Conference. 2004. Albuquerque,
NM: Integrated Petroleum Environmental Consortium, University of Tulsa, Tulsa, OK
USA
42
RPSEA Report number: 11122-31
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
Yang, Y., Y.I. Cho, and A. Fridman, Plasma Discharge in Liquid: Water Treatment and
Applications. 2012, New York: CRC Press.
Czernichowski, A., Gliding arc: Applications to engineering and environment control.
Pure Appl. Chem., 1996. 66: p. 1301-1310.
Kalra, C.S., et al., Gliding arc in tornado using a reverse vortex flow. REVIEW OF
SCIENTIFIC INSTRUMENTS, 2005. 76: p. 025110.
Locke, B.R., R. Burlica, and M.J. Kirkpatrick, Formation of reactive species in gliding
arc discharges with liquid water. Journal of Electrostatics, 2006. 64(Copyright 2006,
IEE): p. 35-43.
Kim, H.-S., et al., Effects of H2O2 and Low pH Produced by Gliding Arc Discharge on
the Inactivation of Escherichia Coli in Water. Plasma Medicine, 2011. 1(3-4).
APHA-AWWA-WPCF., Standard methods for the examination of water and wastewater.
17th ed. 1989: APHA American Public Health Association.
Yang, Y., et al., Effect of a plasma-assisted self-cleaning filter on the performance of PWT
coil for the mitigation of mineral fouling in a heat exchanger. International Journal of
Heat and Mass Transfer, 2010. 53(1): p. 412-422.
(NERL), U.S.E.P.A.N.E.R.L., EPA Method 160.1- Residue, Filterable (Gravimetric,
Dried at 180 oC), Issued 1971, National Environmental Methods Index: nemi.gov.
Bott, T.R., Fouling of heat exchangers. Vol. 26. 1995, New York: Elsevier Science.
Khawaji, A.D., I.K. Kutubkhanah, and J.M. Wie, Advances in seawater desalination
technologies. Desalination, 2008. 221(1-3): p. 47-69.
Mondal, S. and S.R. Wickramasinghe, Produced water treatment by nanofiltration and
reverse osmosis membranes. Journal of Membrane Science, 2008. 322(1): p. 162-170.
Yang, Y., et al., Mechanism of Calcium Ion Precipitation from Hard Water Using Pulsed
Spark Discharges. Plasma Chemistry and Plasma Processing, 2011. 31(1): p. 51-66.
Yang, Y., et al., Application of pulsed spark discharge for calcium carbonate precipitation
in hard water. Water Research, 2010. 44: p. 3659-3668.
Yang, Y., et al., Mineral Fouling Control by Underwater Plasma Discharge in a Heat
Exchanger. ASME Journal of Heat Transfer, 2011. 133(5): p. 054502.
Yang, Y., et al., Removal of CaCO3 scales on a filter membrane using plasma discharge
in water. Int. J. Heat Mass Transfer, 2009. 52(21): p. 4901-4906.
Ranck, J.M., et al., BTEX removal from produced water using surfactant-modified zeolite.
Journal of Environmental Engineering, 2005. 131(3): p. 434-442.
Duff, R., D. Webley, and R. Scott, Solubilization of minerals and related materials by 2ketogluconic acid-producing bacteria. Soil Science, 1963. 95(2): p. 105-114.
Hall, P.A., L. Ledenbach, and R.S. Flowers, Acid-producing microorganisms.
Compendium of Methods for the Microbiological Examination of Foods, 2001. 4: p. 201207.
Barton, L., Sulfate-reducing bacteria. Vol. 8. 1995: Springer.
Amann, R., et al., Molecular and microscopic identification of sulfate-reducing bacteria
in multispecies biofilms. Applied and Environmental Microbiology, 1992. 58(2): p. 614623.
Videla, H.A. and L.K. Herrera, Microbiologically influenced corrosion: looking to the
future. International microbiology, 2005. 8(3): p. 169.
Postgate, J.R., The sulphate reducing bacteria. 1979: CUP Archive.
43