Content uploaded by Roger Gaudreault
Author content
All content in this area was uploaded by Roger Gaudreault on Sep 12, 2016
Content may be subject to copyright.
1
Control of Deposition Risks in High-Silica Boiler Waters:
A Novel Approach Using Purified Tannin Chemistry
Roger Gaudreault1*, Salman Safari1,2, Theo G.M. van de Ven2, Max Junghanns3
1TGWT Clean Technologies Inc., Longueuil, QC, Canada, 2Department of Chemistry, McGill
University, Montreal, QC, Canada, 3Xelera SA de CV, Puebla, Mexico.
*Corresponding author: rgaudreault@tgwt.com
ABSTRACT
The use of make-up water sources having a high-silica concentration has proven to be a major challenge
for the operation of industrial steam boilers. Despite advances in conventional deposit control
technologies, high-silica boiler water (>150 mg/L as SiO2) continues to have a major impact on the
operation and efficiency of boilers. The primary concern is avoiding silica/silicate-based deposits which
are thermally insulating and significantly reduce heat transfer. Conventional treatment requires limitations
on boiler silica levels which can result in operation at low cycles of concentrations, excessively high make-
up water and blowdown rates, and high fuel usage rates.
Here, case studies show that Purified Tannin treatment (a green chemistry process) not only allows boiler
operations at significantly higher silica levels while inhibiting formation of scale, but also shows visual
evidence suggesting removal of previously deposited scale. Moreover, laboratory results using
photometric dispersion analysis (PDA) and dynamic light scattering (DLS) show that silica/tannin are still
stable at silica levels well above any established industry guidelines.
This work provides a new model for steam boilers that utilize high-silica water for their make-up. It
demonstrates that operation at higher cycles of concentration and boiler silica levels is attainable while
controlling deposition risks in the boiler. Finally, operational guidelines are proposed for boilers operating
with Purified Tannin.
KEYWORDS
Green chemistry, tannin molecules, silica, colloids, steam boilers, scale, corrosion inhibition
INTRODUCTION
Control of corrosion and scale is critical for the protection of capital assets throughout industry and society.
It is especially important for the treatment of industrial steam boilers where metal surface temperatures
exceed the boiling point of water at the boiler operational pressure. It is well known that temperature
increases the potential for corrosion of metals and scale deposition. Generally speaking, methods to control
scale and corrosion have been on-going processes in many industries.
Because of economic concerns and climate change, water and energy usage and exhaust gas emissions
have become major factors that must be considered in the overall benefits of water treatment. For industrial
2
steam boilers, not only the cost of the boiler treatment is considered, but we can now calculate the savings
from reducing the use of fuel and water, and the amount of greenhouse gas (GHG) emissions (depending
on the state, province, or country). Interestingly, in the near future it is expected that the industry will be
able to estimate the disability-adjusted life years (DALYs), resulting from water treatment, through life
cycle impact assessment methodologies. Nonetheless, we must calculate as precisely as possible the
following savings: energy, GHG, and water!
One of the difficult aspects of scale and corrosion inhibition in steam boilers is to minimize scale formation
on the tube surface, i.e. below the critical point where heat transfer efficiency is reduced. The industry has
been able to correlate, and in some cases, quantify the additional amount of fuel required for steam
generation per thickness of boiler tube scale. Obviously, this correlation is a function of the type of metal
and the chemical composition of the scale, i.e. organic and/or inorganic (e.g. carbonates, hardness salts,
silica-based, and phosphates). Scaling is not specific to boilers or equipment, it also occurs in plants and
living organisms, e.g. inside daphnia. Scale formation at high pH and temperature is easier to trigger and
normally a challenge to prevent and difficult to remove. Several ways to remove scale can be used, but
may cause challenges and safety issues. Some of the approaches can be performed during normal operation
or only when the boiler and associated equipment are shut down, which can be very expensive. Obviously,
a hybrid approach is possible, for example, an acid cleaning followed by an on-line addition of synthetic
or natural chelating/dispersing agents or polymers.
There is a consensus among industry experts that high-silica concentration is a major challenge for the
operation of steam boilers. Despite advances in conventional deposit control technologies, high-silica
boiler water (>150 mg/L) continues to have a major impact on the operation and efficiency of boilers. The
solubility of crystalline silica (quartz) is quite low with 5-6 mg/L at 25°C and pH < 9, with the solubility
of amorphous silica ranging from 120-150 mg/L at 25°C and pH 8-8.5 [1]. Colloidal silica that enters
equipment with the feedwater can settle on the boiler tubes in form of silicates [2]. Silica-based deposits
are highly insulating and greatly reduce heat transfer. For instance, deposits composed of silicates were
found to be responsible for the failures of boiler tubes in thermal power plants and it is also a great problem
in geothermal energy utilization [3]. It is widely accepted that the polymerization of silica monomers is
the formation mechanism of amorphous silica deposits [4,5]. Conventional treatment requires limitations
on boiler silica levels which can result in operation at low cycles of concentrations, excessively high make-
up water and blowdown rates, and high fuel usage rates. Hence, the control of deposition risks in high-
silica boiler water is a challenging task that has not yet been fully addressed.
This work describes case studies to support the possibility to control deposition risks in high-silica boiler
water using Purified Tannin chemistry. The second part of this work conveys scientific evidence that
silica/tannin are still stable at silica levels well above any established industry guidelines. Finally, we will
propose operational guidelines for protecting steam boilers when using green chemistry through tannin
treatment and compare them with ASME guidelines.
MATERIALS AND METHOD
Chemical and Solutions
The Purified Tannin solution was prepared by diluting concentrated TG-3106, provided by TGWT Clean
Technologies Inc., using Reversed Osmosis (RO) water. Pure colloidal silica nanoparticles (30 nm, 50%
3
(wt/wt) Ludox TM-50) were purchased from Sigma Aldrich, Ontario, Canada. Nissan Chemical America
kindly gifted a sample of aqueous colloidal suspension of 200 nm pure silica nanoparticles (40% (wt/wt)
MP-2040). The size of silica nanoparticles was measured with DLS, and their diameters were found to be
35±2 and 207±2 nm, with a polydispersity (PD) index of 0.19 and 0.03, respectively. Ferrous sulfate,
ferric sulfate, sodium sulfate, sodium chloride, magnesium chloride, and calcium chloride were supplied
by Sigma Aldrich, Ontario, Canada. For pH adjustment, concentrated sodium hydroxide and sulfuric acid
from AquaPhoenix, Pennsylvania, USA, were used.
Photometric Dispersion Analyzer (PDA)
Changes in the stability of silica and/or Purified Tannin, hereinafter referred to as tannin, were monitored
with a Photometric Dispersion Analyzer (PDA 2000 Rank Brothers, Cambridge, UK) [6,7]. The
silica/tannin suspension was pumped through a transparent 3 mm diameter tubing into the photocell of the
PDA, which monitors the fluctuations in intensity of transmitted light (figure 1). In the present study, the
suspension was pumped using a peristaltic pump at a constant flow rate corresponding to an average shear
rate (G) of ~200 s−1. This experimental setup allows measurements of the aggregation and break-up
kinetics.
Figure 1: Experimental set-up for the Photometric Dispersion Analyzer (PDA).
The transmitted light intensity has two major components:
V
which represents the average transmitted
light intensity (DC; direct current) and a much smaller component (
rms
V
) defined as the standard deviation
around the mean signal. The
rms
V
increases considerably as particles aggregate, whereas the aggregation
only slightly changes
V
. Therefore, a significant increase in the ratio (R), also referred to as aggregation
index (AI), is a clear indication of particle aggregation, with R defined as:
V
V
Rrms
Eq. 1
For dilute systems, R varies linearly with the square root of the concentration, and for polydisperse systems
R is approximately linear with the particle radius, a [6,7]:
4
aCNR2/1
Eq. 2
where N is the number of particles per unit volume, and C is a constant. Changes in R (ΔR) correlate with
changes in size (Δa) measured by static light scattering/diffraction [8]. PDA is a powerful technique to
monitor particles stability at various shear rates in real time [9-13]. Interestingly, the PDA detection limit
of about 0.5 micron is about 100 times smaller than what a human eye can actually see.
Dynamic Light Scattering (DLS)
Particle size analyses were performed using a Brookhaven 90Plus particle size analyzer. DLS results are
average of three measurements without stirring at 25 ± 0.1 °C. Measurements were carried out at an angle
of 90° from the incident beam.
RESULTS AND DISCUSSION
This work is divided into three parts. The first part is to present case studies showing the use of tannin for
the control of deposition risks in high-silica boiler waters. The second part is to present some scientific
evidence for the stability of silica/tannin at conditions well above industry guidelines for boiler water
operation. Finally, operational guidelines are proposed for protecting steam boilers when using tannin, and
those guidelines are compared to ASME guidelines.
Case Studies
Three steam boiler case studies are presented where the source for make-up was high-silica water. In all
cases, conventional treatment technology was replaced with tannin technology, and the boiler cycles of
concentration were significantly increased, hence boiler silica content were much higher than traditional
limits [14]. A few critical boiler water parameters were plotted to show the trends before and after tannin
treatment was implemented. The average and standard deviation were calculated for the pH, silica
concentration, un-neutralized (UN) and neutralized (N) conductivities, M-Alkalinity, and residual tannin.
The data of the transition period between treatments were excluded to prevent any skewed average.
Pictures of boiler tubes, taken during openings/inspections, are provided to show visual depiction of the
boiler conditions before and after tannin treatment.
Poultry Industry Feedmill (USA)
A poultry industry feedmill was using a 200 HP fire-tube boiler for process and heating applications. The
boiler was operating at approximately 100 psig and burning propane (LPG) for fuel. The mill called for
about 30,000,000 lbs per year of steam and returned about 20% of that to the boiler as condensate. The
well water used as the source of make-up for the boiler contained high levels of silica that ranged
seasonally between 25 and 40 ppm as SiO2. Along with the low percentage condensate return, the high-
silica in the make-up limited the boiler operation to about 15 cycles of concentration. A conventional
sulfite, polymer, caustic, and amine program was used to treat the boiler system. At those 15 cycles, as
depicted in figure 2, some deposition was noticed at annual boiler inspections and the boiler silica levels
climbed above 150 ppm (table 1).
5
Figure 2: 2014 inspection of top (left and middle) and bottom (right) tubes while using
conventional treatment.
After the inspection in late November of 2014, the boiler treatment was switched to a tannin program on
December 2nd. The conductivity control range was increased significantly while using the same make-up
water source. Table 1 summarizes the boiler operational data while using conventional treatment up to 12-
01-2014 and while using the tannin treatment from December 2nd 2014 to the date of the writing of this
paper (June 2016). It can be seen that the conductivity, M-alkalinity, and silica levels in the boiler were
maintained well above established industry guidelines while using tannin.
Table 1: Boiler water parameters before and after using tannin.
Boiler water parameter
Conventional treatment
Tannin treatment
Cycles of concentration*
15 (10)
50 (35)
% Blowdown
6.4 (10)
1.9 (2.9)
pH
10.5 ± 0.3
11.7 ± 0.3
Silica as SiO2 (ppm)
179 ± 80
1,288 ± 235
Un-Neutralized conductivity (µmhos cm-1)
2,348 ± 403
10,720 ± 1026
M-alkalinity as CaCO3 (ppm)
355 ± 129
1,497 ± 605
Residual tannin (ppm at 420 nm)
Not Applicable
171 ± 60
*Based on Un-Neutralized conductivity and (Neutralized conductivity).
Several benefits of switching to tannin treatment and operation at significantly higher cycles are reduction
in boiler blowdown and sewer water discharge, make-up water requirements, propane usage for fuel, and
level of greenhouse gas (GHG) emissions. These results were tabulated and documented as boiler house
budgetary savings for the feedmill. Figure 3 depicts the condition of the boiler in November of 2015, after
one year of operation using tannin treatment. The UN conductivity control set-point was typically 12,000
µmhos cm-1 and the typical boiler silica level was 1,200 ppm as SiO2. It can be seen in figure 3 that pre-
existing scale on the tubes seen in figure 2 has been partially removed.
In conclusion, the poultry industry feedmill boiler was allowed to safely operate at much higher
conductivity and cycles of concentration, and at extremely high-silica levels. The boiler inspection
revealed that the tannin treatment not only inhibited scale formation, but also showed visual evidence of
scale removal.
6
Figure 3: 2015 inspection of top (left and middle) and bottom tubes while using tannin treatment.
Figures 4 to 12 are presented to show the key data trends for the feedwater and boiler water for the periods
before and after the tannin treatment. Charts depict raw data trends and calculated boiler cycles going back
several years of operation to current date. To minimize space in this paper, certain feed water and boiler
water data sets were plotted together. To emphasize important points during the operation using tannin
treatment, data sets starting in December of 2014, were isolated and plotted.
Figure 4: Boiler feedwater and boiler water pH versus time (tannin treatment started 12/02/2014).
7
Figure 5: Boiler feedwater silica levels versus time (tannin started 12/02/2014).
Figure 6: Boiler water silica levels versus time (tannin started 12/02/2014).
8
Figure 7: Boiler feedwater and boiler water (Un-Neutralized) conductivity versus time
(condensate conductivity inset).
Figure 8: Boiler water Neutralized conductivity versus time.
9
Figure 9: Boiler cycles versus time (based on Un-Neutralized and Neutralized conductivities).
Figure 10: Boiler cycles versus time (based on silica and Neutralized conductivity).
10
Figure 11: Boiler water M-alkalinity versus time.
Figure 12: Boiler water residual tannin versus time (using Hach DR 890 Unit at 420 nm
wavelength).
11
Textile Industry (Mexico)
Mexico is known for having volcanic formations which result in many regions of the country having
ground water containing significant levels of silica. This has caused silica control to be a major issue for
steam boiler operation throughout the country. Silica levels can range from 40 to 105 ppm in ground water.
A textile manufacturer in Mexico has been dealing with this issue for many years, and utilized make-up
water containing 60 – 80 ppm of silica as SiO2.
The facility has a 3 fire-tube boilers burning natural gas for fuel: a 200 HP and a 400 HP operating since
1998, and a 500 HP operating since October of 2014. The operating pressures for the boilers range from
87 to 116 psig, and the condensate return for all 3 boilers was approximately 33%. The low condensate
return rate and high-silica levels of the make-up water significantly limited the number of boiler cycles of
concentration. Even at low cycles, boiler scale was encountered requiring periodic acid cleanings.
The boilers were treated with a conventional phosphate, caustic, and silica dispersant program until June
2015. The low cycles of operation resulted in a high blowdown rate, and a huge requirement for make-up
water and natural gas for fuel. On June 8th 2015, the boiler treatment program was switched to tannin to
address the deposit issues and to allow operation at higher cycles of concentration.
Under the tannin program, pH, silica, conductivity, residual tannins, chlorides, and the boiler cycles of
concentration were initially monitored on a weekly basis, followed by monitoring on a monthly basis.
Typical values for all data monitored are listed in table 2. Caustic feed for boiler alkalinity control was not
required with the tannin program. Although the tannin program does not require measurement of either P,
M or OH alkalinity, the M-alkalinity was measured to provide comparison with conventional treatment.
Table 2: Boiler water parameters before and after tannin treatment.
Boiler 200 HP
Boiler 500 HP
Boiler water parameters
Conventional
treatment
Tannin
Conventional
treatment
Tannin
Cycles of concentration*
5
13
5
16
% Blowdown
20
7.7
20
6.1
pH
11.0 ± 0.4
11.4 ± 0.3
11.3 ± 0.5
11.6 ± 0.4
Silica as SiO2 (ppm)
297 ± 64
748 ± 165
289 ± 47
946 ± 304
Un-Neutralized conductivity
(µmhos cm-1)
4,227 ± 768
8,184 ±
1,775
4,710 ± 1,525
10,331 ±
3,079
M-Alkalinity as CaCO3
(ppm)
1,063 ± 367
2,338 ± 652
1,275 ± 438
2,957 ± 953
Residual tannins
(ppm)
Not Applicable
139 ± 80
Not Applicable
203 ± 101
*Number of cycles based on chlorides.
As a result of switching to the tannin program, the number of boiler cycles was increased from 5 to 13
(200 HP) and from 5 to 16 (500 HP). The % blowdown was reduced from 20% to 7.7% (200 HP) and 20%
to 6.1% (500 HP). These higher boiler cycles resulted in significant reductions in water and natural gas
requirements. The Un-Neutralized conductivity control set-point ranged from 8,000 to 10,000
12
µmhos cm-1 for the 2 boilers. The typical boiler silica levels averaged 750 ppm and 950 ppm as SiO2,
which are well above industry guidelines for fire-tube boilers operating at <300 psig.
Figure 13 shows the condition of the 200 HP boiler using conventional treatment at low cycles in
comparison to that after one full year of operation using the tannin treatment at much higher cycles. On-
line clean-up of old deposits was noted during inspection. Figure 14 shows the condition of the newer 500
HP boiler six months after using the tannin treatment at higher cycles. The boiler tubes are noticeably
clean.
In conclusion, the textile manufacturer’s boilers were allowed to safely operate at much higher
conductivity and cycles of concentration, and at extremely high-silica levels. The boiler inspections
revealed that the tannin treatment not only inhibited scale formation, but also showed visual evidence of
scale removal for the 200 HP boiler.
Figure 13: 200 HP boiler, December 2014, conventional treatment (left); June 2016, after one year
of tannin treatment (right).
Figure 14: Inspection of 500 HP boiler, December 2015, after 6 months of tannin treatment.
13
Figures 15 to19 are presented to show the key data trends for the boiler water for the periods before and
after the tannin treatment. Charts depict raw data trends going back several years of operation to current
date.
Figure 15: Boiler water pH versus time (tannin started June 8, 2015).
Figure 16: Boiler water silica levels versus time (tannin started June 8, 2015).
14
Figure 17: Boiler water Un-Neutralized conductivity versus time.
Figure 18: Boiler water M-alkalinity versus time.
15
Figure 19: Boiler water residual tannin versus time.
Textile Industry (USA)
A global textile company was using four fire-tube boilers ranging 300 HP to 750 HP to generate
approximately 100,000,000 lbs per year of steam for various process and heating applications. The boilers
operated at 125 psig while burning natural gas for fuel, and the system returned about 30% of the steam
as condensate. In the mid 2000’s, the plant had been using well water for make-up that contained silica
levels ranging from 15 to 25 ppm as SiO2. Although limiting boiler cycles of concentration to a low level,
the conventional treatment program in use failed to maintain clean boiler tubes. The boilers were scaled
with heavy Ca, Mg, and Silicate deposits. In 2006, a decision was made to attempt to clean up the boilers
on-line. The make-up source was switched to city water which was more expensive but contained
significantly less silica, and the boiler treatment was switched to a chelant based program. The boiler
cycles of concentration remained low, with an Un-Neutralized conductivity control point of 1,600 µmhos
cm-1, and a 7% blowdown rate. After 2 full years of running, the boiler make-up water costs were hurting
the budget, and the on-line cleaning was deemed to be too slow and unsuccessful. Figure 20 depicts the
condition of the boiler tubes after two years of trying to clean up the scale with an on-line, chelant
treatment.
16
Figure 20: 2008 boiler tubes inspection after 2 years of on-line chelant clean up.
In late 2008, the boiler treatment was switched to a Purified Tannin program to speed up the on-line
cleaning. From day 1, the make-up source was switched back to the well water containing high silica
levels. The boiler conductivity control point was initially raised to 6,000 µmhos, yielding higher cycles of
concentration and a much lower blowdown rate of 2.5%. In early 2009, two of the boilers were opened for
inspection after 4 months of Purified Tannin treatment to evaluate conditions. The engineering and
maintenance teams were impressed with the cleaning progress, especially at the higher boiler cycles and
silica levels. Figure 21 depicts the condition of the boiler tubes after just 4 months of operation where
some bare tube surface was already exposed.
Figure 21: 2009 boiler tubes inspection after 4 months of Purified Tannin treatment.
With the positive results from the inspection and the blessing of the plant personnel, the conductivity
control point for the boilers was raised to 10,000 µmhos for the duration of the on-line cleaning process.
The Purified Tannin treatment cleaned up the boilers to satisfactory levels and continues to maintain the
boilers clean while being on-line for approximately 8 years. The conductivity control point since late 2008
has ranged from 8,000 to 12,000 µmhos, while the boiler silica levels and M-alkalinity have ranged from
450 to 850 ppm as SiO2 and from 690 to 2700 ppm as CaCO3, respectively (table 3).
The use of well water versus city water for boiler make-up represented significant savings for the plant.
The increased boiler cycles further reduced the cost of make-up water as the blowdown rate was reduced
from 7.1% to 2.3% (table 3). The plant also realized savings on their sewer bill, as less water was being
discharged to the city sewer. With clean boiler tubes and boiler operation at much higher cycles of
concentration, the requirement for natural gas to fuel the boilers was also reduced, resulting in measured
savings on gas purchases.
17
Table 3: Textile Industry: Comparison between Conventional and Purified Tannin treatment.
Water treatment program
Conventional
treatment
Purified
Tannin
treatment
Boiler
operation
Cycles of
concentration*
14
44
% Blowdown
7.1
2.3
Boiler
water
quality
pH
11.8
12.3
UN-Conductivity
(µmhos cm-1)
1,600
10,480
Silica as SiO2
(ppm)
66
706
Type of Feedwater
City Water
Well Water
*Based on UN-conductivity.
Stability of Tannin and Silica (PDA characterization, with shear)
The goal of this laboratory investigation is to shed light on the stability of purified tannin in high-silica
boiler water. Tannins are macromolecules having their size in the colloidal range that do not settle. Because
they experience about 1021 collisions per second with water molecules [15], they are constantly moving
randomly. Moreover, these macromolecules have functional groups that become highly negatively charged
at high pH, which keeps them stable in solution [16,17]. Silica also becomes more negatively charged as
pH increases [18-20], hence they both behave similarly, e.g. they are stable at higher pH.
Silica solubility drastically increases at pH above ~10 [21,22]. Then, polymerized silica and silicate salts
of metal ions or other cations, can be formed, which can trigger the nucleation of crystals in the bulk
solution.
There are two possible mechanisms to induce scaling: 1) scale can be either directly induced on metal
surfaces (e.g. boiler tubes); and 2) indirectly in the bulk of process water and subsequently depositing on
the tubes. We performed PDA experiments to gather useful information on bulk nucleation. In other words,
the destabilization/aggregation of silica/silicate/tannin in solution gives information about the propensity
of precipitation in solution, which can subsequently deposit on the boiler metal surfaces. Thus, the
following investigation is an attempt to get insights on how to minimize the propensity of high-silica level
to scale boilers.
Tannin
Tannin solutions, without silica, are stable in all investigated conditions, i.e. pH ranges from 7 to 12, NaCl
concentration up to 0.15 M (14,200 µmhos cm-1), conductivity up to ~50,000 µmhos cm-1 (0.5 M Na2SO4),
50 ppm Fe2+/3+ as Fe, 100 ppm Ca2+ as Ca, and 50 ppm Mg2+as Mg (not shown here).
18
Silica 30 and 200 nm
Table 4 shows the stability of colloidal 30 nm silica nanoparticles (up to 5000 ppm), without tannins, in
the pH range of 7 to 12 and NaCl concentrations up to 0.15 M (14,200 µmhos cm-1) [19]. Then, additional
experiments were performed with larger 200 nm SiO2 nanoparticles. The 200 nm silica nanoparticles
behave differently than the 30 nm SiO2 nanoparticles, the former having a faster aggregation rate and lower
dissolution rate. In other words, at pH 12, the 30 nm nanoparticles dissolve faster, while the 200 nm
nanoparticles aggregate before they are completely dissolved. In any case, aggregates that are formed are
temporary and easily disappear when subjected to shear/agitation (table 4).
Tannin and silica
Similar results were observed for a mixture of tannin and silica nanoparticles under the same conditions.
Under most conditions, the tannin and silica nanoparticles are stable (table 4). These findings are in
agreement with the chemical properties of both tannin and silica nanoparticles, i.e. having a highly negative
charge density at higher pH, which enhances their stability through electrostatic repulsion. However, at
extremely high conductivity, e.g. ~50,000 µmhos cm-1 (0.5 M Na2SO4), 5000 ppm colloidal silica (30 nm)
starts to aggregate at pH ~10 and as pH increases aggregation tends to occur at lower silica concentration,
e.g. 300 ppm at pH~12. Those aggregates are permanent since they do not break up under experimental
shear of ~200 s-1 (Tables 4 and 5).
Table 4: Silica and tannin/silica (30 and 200 nm) stability (PDA characterization, with shear).
Parameters
30 nm Silica
200 nm Silica
Silica up to
5000 ppm
Tannin (275 ppm),
silica as SiO2
(up to 5000 ppm)
Silica as SiO2
(up to 5000 ppm)
Tannin (275 ppm),
silica as SiO2
(up to 5000 ppm)
pH 7-12
stable
Stable
stable
stable
Cond. up to 14,200
µmhos cm-1
(0.15 M NaCl)*
stable
Stable
stable
stable
Fe2+ as Fe; up to ppm
stable
Stable
stable
stable
Fe3+ as Fe; up to ppm
stable
Stable
stable
stable
Ca2+ as Ca;
up to 100 ppm
temporary
aggregates1
permanent
aggregates2
stable up to 150
ppm of SiO2
stable up to 500 ppm
of SiO2
Mg2+ as Mg;
Up to 50 ppm3
temporary
aggregates1
temporary
aggregates1
stable up to 150
ppm of SiO2
stable up to 150 ppm
of SiO2
Cond. 50,000 µmhos
cm-1, pH 12
(0.5 M Na2SO4)
permanent
aggregates3
permanent
aggregates3
N.D.
N.D.
1 Temporary aggregates are reversible and disappear after few hours or days.
2 Up to 70 ppm of Ca2+ (as Ca2+) and pH 7-12 stable aggregates below PDA detection limit, 80 ppm and
pH>11 permanent aggregates.
3 Up to 300 ppm of SiO2 no aggregates. Higher silica concentration aggregates start to appear.
N.D.: Not determined.
19
Tannin, silica and calcium ions
The effect of calcium on the stability of silica was investigated and the results show the presence of
temporary settled aggregates. These temporary aggregates could be due to the bridging of partially
dissolved outer layers of silica nanoparticles by calcium ions. They easily break up because the bridges
might be weakened as more silica gets solubilized, resulting in a distribution of Ca2+ ions over more
partially solubilized chains [18]. On the other hand, figure 22 shows that permanent settled aggregates of
silica 30 nm/tannin are induced at calcium concentrations equal or greater than 80 ppm (as Ca2+ or 200
ppm CaCO3). Moreover, figure 23 shows that 70 ppm of calcium (pH 7 to 12) does not induce aggregates,
i.e. the baseline of the aggregation index (AI) does not change (larger AI corresponds to the presence of
bigger aggregates). Interestingly, 80 ppm calcium ions (as Ca2+) initiate fast aggregation of silica
nanoparticles, reversible up to pH 11 (about 8100 seconds), but irreversible at pH>11.
Figure 22: Effect of calcium on tannin/silica stability: 25, 50 and 80 ppm Ca2+ as Ca,
5000 ppm SiO2 (30 nm), pH~12.
However, the results from 200 nm SiO2 differ, wherein no aggregation occurs at silica levels of 150 ppm
or less, while at a higher silica concentration aggregates start to appear. In the presence of tannin and at
100 ppm Ca2+, tannin improves the stability limit from 150 to 500 ppm of silica (table 4).
Ca2+: 25 ppm
Ca2+: 50 ppm
Ca2+: 80 ppm
20
Figure 23: PDA Aggregation Index of silica/tannin suspension: 275 ppm Purified Tannin and 5000
ppm SiO2 (30 nm) in the presence of Ca2+ ions, shear rate (G) of ~200 s-1. Note the results were
truncated to after 5000 seconds to accommodate eventful results.
Tannin, silica, iron, and magnesium ions
Both tannin and silica nanoparticles maintained their stability in the pH range 7 to 12, up to 50 ppm of
Fe2+ and Fe3+ as Fe, most likely due to the precipitation of ferrous and ferric hydroxide at pHs higher than
8 (table 4). On the other hand, magnesium (50 ppm as Mg2+ or 173 pm as MgCO3, pH 7 to 12) triggered
temporary destabilization of silica nanoparticles at higher pHs, which leads to the formation of temporary
aggregates (tables 4 and 5). The observed aggregates disappear after few hours, i.e. the time needed that
silica nanoparticles fully dissolve.
21
Table 5: Summary of stability of tannin and silica (30 nm).
Type of system
/aggregation
Concentra-
tion*
(ppm)
Aggregates
Photos of
solution/suspension
Tannin
275
NO
Silica
5000
NO
Tannin-silica
275-5000
NO
Tannin-calcium
275-100
NO
Silica-calcium
5000-80
Temporary
(left)/ few
hours later
(right)
Tannin-calcium-silica
275-80-5000
Permanent
Tannin-Fe2+/Fe3+-silica
275-50/50-
5000
NO
Tannin-Mg2+-silica
275-25/50-
5000
NO at 25
ppm/tempora
ry at 50 ppm
Tannin-silica-Na2SO4
(0.5 M), cond. of
50,000 µmhos cm-1
275-5000
Permanent
*Silica as SiO2
22
Particle size (DLS characterization, without shear)
The next series of experiments were performed to determine the particle size of SiO2, Purified Tannin and
a mixture of SiO2 and tannin, hence to validate whether tannin shows dispersive properties. Figure 24
shows that silica particles at 5000 ppm concentration have a bimodal size distribution, in which one peak,
~25 nm, corresponds to individual nanoparticles and the other, ~80 nm, corresponds to doublets or triplets
(concentration effect) of nanoparticles. As previously mentioned, the average size of silica nanoparticles
is 35±2nm. Purified Tannin, on the other hand, exhibits a broad size distribution from 80 nm to 1200 nm
probably due to the association of tannin macromolecules [23]. Interestingly, mixing Purified Tannin and
SiO2 results in a size distribution similar to that of SiO2, because silica particles outnumber tannin
macromolecules in the mixture and their stability is not influenced by the presence of tannin.
Figure 24: Size distribution of 30 nm SiO2, Purified Tannin, and mixed suspension of 5000 ppm of
30 nm SiO2 and 275 ppm of Purified Tannin. DLS results attest to the presence of silica doublets and
triplets.
The effect of calcium ions was also studied and as seen in figure 25(a) silica nanoparticles aggregate up to
70 ppm of calcium ions after which the aggregates start to settle within few minutes. Similarly, Purified
Tannin shows an increase in its size with addition of calcium; however, it maintains its stability (figure
25(b)). Interestingly, by comparing figures 25(a) and (c) one can notice dispersive effect of Purified Tannin
on silica nanoparticles aggregates which leads to break-up of aggregates into smaller entities, which do
not settle and remain stable in the bulk suspension as confirmed by PDA method. Results when using
magnesium instead of calcium showed qualitatively similar effect.
23
Figure 25: Effect of calcium on size distribution of (a) 30 nm SiO2, (b) Purified Tannin (c) mixture
of 5000 ppm of 30 nm SiO2 and 275 ppm of Purified Tannin at pH~12.
(a)
(b)
(c)
24
Figure 26: Optical microscopy (dry state): left) pure silica 5000 ppm as SiO2 (30 nm); and right) 275
ppm Purified Tannin, 80 ppm calcium as Ca2+, and 5000 ppm of silica as SiO2 (30 nm).
Optical microscopy
Preliminary optical microscopy images show typical striped patterns for dried colloidal silica [24] and
aggregates of tannin/silica-silicates in the presence of 80 ppm of Ca2+ (figure 26). Further investigation is
under way to obtain a better understanding of these images.
Finally, although this laboratory investigation supports that tannin/silica suspensions are stable at silica
levels well above any established industry guidelines, additional research is needed to elucidate the
underlying mechanisms.
Boilers operational guidelines
Based on case studies presented in this work, supported by more than 10 years of industrial field experience
in North America, as well as scientific evidence from previous publications and this work, operational
guidelines are proposed. Consequently, tables 6 and 7 show the ASME guidelines as well as the proposed
guidelines for boilers operating with Purified Tannin. These are conservative guidelines since Purified
Tannin can perform at higher silica and conductivity shown in table 7.
Table 6: Conventional treatment: ASME guidelines for modern
industrial fire-tube boilers for reliable continuous operation [14].
Boiler water
Drum pressure
(psig)
Silica
(ppm as SiO2)
Total alkalinity
(ppm CaCO3)
Conductivity
(µmhos cm-1) (Un-
neutralized)
0 ̶ 300
≤ 150
<700
<7,000
These apply when total hardness <1 ppm and oxygen concentration <7 ppb.
25
Table 7: Purified Tannin treatment: Proposed guidelines for modern industrial
fire tube and water tube boilers for reliable continuous operation.
Boiler water
Drum
pressure
(psig)
pH
Silica
(ppm as
SiO2)
Conductivity*
(µmhos cm-1)
(un-neutralized)
Residual tannins
(ppm at 420 nm)
0 ̶ 300
11.5±1.0
≤ 750
<10,000
175±25
*Function of boiler blowdown rate. Targeted residual tannin for scaled boilers is
225±25 ppm (measured at 420 nm). These apply when total hardness <1 ppm and
oxygen concentration <7 ppb.
CONCLUSIONS
Here, case studies show that a green chemistry treatment with Purified Tannin not only allows boiler
operations at significantly higher silica levels while inhibiting formation of scale, but also shows visual
evidence of removal of previously deposited scale. Laboratory results show that silica/tannin are still stable
at silica levels well above any established industry guidelines. Laboratory studies also show that Purified
Tannin has a dispersive effect on silica nanoparticles aggregates both in the absence and presence of ions
such as Ca, Mg and Fe, which minimizes the propensity of precipitation in the bulk solution, and
subsequently decreases the deposition risks on the boiler metal surfaces. Several benefits of switching to
green chemistry through tannin treatment and operating at significantly higher cycles are: a reduction in
boiler blowdown and sewer water discharge, make-up water requirements, propane usage for fuel, and
level of greenhouse gas (GHG) emissions. Finally, guidelines for boilers operating with Purified Tannin
are proposed.
ACKNOWLEDGEMENTS
First, we would like to dedicate the USA Textile Industry case study to late Mr. Ritchie Jenkins,
plant engineer, who passed away in 2016.
The authors gratefully acknowledge the financial support from Mitacs Accelerate Program. The authors
thank Louis-Philippe Cloutier, Dave Ritz and Louis Dubois from TGWT Clean Technologies Inc., and
Rafael Gonzalez from Xelera Inc., NC, USA, for valuable input for the case studies. The authors also
acknowledge Christian Fowelin from Korn GmbH for valuable comments, and Janice Ritz from TGWT
Clean Technologies Inc. for proofreading and graphics editing of the manuscript. Finally, we gratefully
thank Condor Technologies, A Division of Azure Water Services, Xelera Inc., and Experienced Water
Solutions Inc. for providing data sets and samples for the industrial case studies.
26
REFERENCES
[1] Hermossila, D., Ordonez, R., Blanco, L., de la Fuente, E., Blanco, A., pH and Particle Structure
Effects on Silica Removal by Coagulation, Chem. Eng. Technol., 2012, 35, 1632-1640.
[2] Amjad, Z., Demadis, K.D., Mineral Scales and Deposits, 2015.
[3] Gill, J. S., Inhibition of silica-silica deposit in industrial waters, Colloids Surf. A Physicochem. Eng.
Asp. 1993, 74, 101-106.
[4] Anthony, A., Low, J.H., gray, S., Childress, A.E., Le-Clech, P., Leslie, G., Scale Formation and
Control in High Pressure Membrane Water Treatment Systems: A review, J. Membr. Sci. 2011, 383,
1-16.
[5] Nishida, A., Shimada, Y., Saito, T., Okaue, Y., Yokoyama, T., Effect of aluminum on the deposition
of silica scale in cooling water systems, J. Colloid Interface Sci., 2009, 335, 18-23.
[6] Gregory, J.; Turbidity Fluctuations in Flowing Suspensions. J. Colloid Interface Sci. 1985, 105, 357-
371.
[7] Gregory, J.; Monitoring Particle Aggregation Processes. Adv. Colloid Interface Sci. 2009, 147−148,
109-123.
[8] Gregory, J.; Nelson, D.W.; Monitoring of Aggregates in Flowing Suspensions. Colloids Surf. 1986
18, 175-188.
[9] Gaudreault, R.; van de Ven, T. G. M.; Whitehead, M. A.; Mechanisms of Flocculation with
Poly(ethylene oxide) and Novel Cofactors, Colloids Surf. A Physicochem. Eng. Asp. 2005, 268: p.
131-146.
[10] Gaudreault, R.; Whitehead, M. A.; van de Ven, T. G. M.; Mechanisms of Flocculation of
Microcrystalline Cellulose by Polyethylene Oxide and Cofactor Corilagin, 13th Fundamental
Research Symposium, Cambridge, September 2005, 1269-1292.
[11] Gaudreault, R.; Di Cesare, N.; Weitz, D.A.; van de Ven, T. G. M.; Flocculation Kinetics of
Precipitated Calcium Carbonate. Colloids Surf. A Physicochem. Eng. Asp. 2009, 340, 56-65.
[12] Sheikhi, A.; Safari, S.; Yang, H.; van de Ven; T. G. M.; Copper Removal Using Electrosterically
Stabilized Nanocrystalline Cellulose, ACS Appl. Mater. Interfaces 2015, 7, 11301-11308.
27
[13] Gaudreault, R.; Di Cesare, N.; van de Ven, T.G.M.; Weitz, D.A.; The Structure and Strength of Flocs
of Precipitated Calcium Carbonate Induced by Various Polymers used in Papermaking, Ind. Eng.
Chem. Res. 2015, 54, 6234−6246.
[14] ASME Guidelines for Modern Industrial Fire-tube Boilers for Reliable Continuous Operation
(http://purewatergroup.com/site_english/img/tabellen/tabel2_4.pdf )
[15] van de Ven, T. G. M. Colloidal Hydrodynamics, Academic Press, 1989, Chapter 2, p. 51.
[16] Gaudreault, R.; Dargahi, M.; Weckman, N.; Olsson, A. L. J.; Omanovic, S.; Schwartz, G.; Tufenkji,
N.; Green Chemistry – with a Special Emphasis on Tannin Molecules for the Protection of
Aluminum Boilers, AWT 2013 Annual Convention and Exposition, October 30-November 1.
[17] Dargahi, M.; Olsson, A. L. J.; Tufenkji, N.; Gaudreault, R.; Green Technology: Tannin-Based
Corrosion Inhibitor for Protection of Mild Steel. Corrosion 2015, 71, 1321-1329.
[18] Kobayashi, M.; Juillerat, F.; Galletto, P.; Bowen, P.; Borkovec, M.; Aggregation and Charging of
Colloidal Silica Particles: Effect of Particle Size, Langmuir 2005, 21, 5761-5769.
[19] Orts-Gil, G.; Natte, K.; Drescher, D.; Bresch, H.; Mantion, A.; Kneipp, J.; Osterle, W.;
Characterization of Silica Nanoparticles Prior to In vitro Studies: From Primary Particles to
Agglomerates. J. Nanopar. Res. 2011, 13, 1593-1604.
[20] Meissner, J.; Prause, A.; Bharti, B.; Findenegg, G. H.; Characterization of Protein Adsorption onto
Silica Nanoparticles: Influence of pH and Ionic Strength, Colloid Polym. Sci. 2015, 293, 3381–3391.
[21] Alexander, G. B.; Heston, W. M.; Iler, R. K.; The Solubility of Amorphous Silica in Water, J. Phys.
Chem. 1954, 58, 453-455.
[22] Amjad, Z., The Science and Technology of Industrial Water Treatment, CRC Press, 2010.
[23] Zanchi, D.; Vernhet, A.; Poncet-Legrand, C.; Cartalade, D.; Tribet, C.; Schweins, R.; Cabane, B.;
Colloidal Dispersions of Tannins in Water-Ethanol Solutions, Langmuir 2007, 23, 9949-9959.
[24] Giorgiutti-Dauphiné, F.; Pauchard, L., Striped Patterns Induced by Delamination of Drying
Colloidal Films, Soft Matter, 2015, 11, 1397-1402.
