Note: Descriptions are shown in the official language in which they were submitted.
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
METHOD OF MONITORING TREATING AGENT RESIDUALS AND
CONTROLLING TREATING AGENT DOSAGE IN WATER TREATMENT
PROCESSES
TECHNICAL FIELD
This invention relates to water treatment. More particularly, this invention
concerns methods of using fluorescent tracers to monitor the residual
concentration of
treating agents in treated water, to determine an optimal water treatment
agent dosage
and to automatically re set the optimal treating agent dosage as necessary to
account
for fluctuations in the characteristics of the treated water.
BACKGROUND OF THE INVENTION
Water, in the course of its use in industrial, municipal and agricultural
applications may be treated with an astounding array of treatment agents
including, for
example, chemicals that enhance solid-liquid separation, membrane separation
process
performance enhancers, antiscalants and anticorrosives that retard or prevent
corrosion
or scale formation and deposition on surfaces in contact with the treated
water,
antifoulants that retard or prevent membrane fouling, biodispersants,
microbial-growth
inhibiting agents such as biocides and cleaning chemicals that remove deposits
from
surfaces that contact the treated water.
Control of treating agent dosage is of paramount iniportance in virtually all
water treatment processes. Obviously, a minimum effective amount of treating
agent
must be maintained in the water for the treatment to have its desired effect.
Conversely, overdosing the treating agent would be at best uneconomical and at
worst
could result in damage to the process or the processing equipment,
particularly in the
case of processes involving the use of membranes as described herein.
Accordingly,
there is an ongoing need for the development of improved methods of monitoring
and
controlling the concentration of water treatment agents in process water.
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
2
SUMMARY OF THE INVENTION
This invention is a method of monitoring residual treating agent and
determining optimal treating agent dosage in process water treated with the
treating
agent comprising sequentially
i) adding a first dose of the treating agent traced or tagged with a
fluorescent
tracer to the process water,
ii) measuring the fluorescence intensity of the process water,
iii) adding a second dose of the treating agent traced or tagged with the
fluorescent
tracer the process water;
iv) measuring the fluorescence intensity of the process water; and
v-a) correlating the change in measured fluorescence intensity of the process
water
at the first and second treating agent doses to the residual concentration of
the treating
agent; or
v-b) correlating the change in the measured fluorescence intensity of the
process
water at the first and second treating agent doses with a non-proportional
change in
measured fluorescence intensity, wherein the non-proportional change in
measured
fluorescence intensity of the process water is used to determine a set point
corresponding to the optimal treating agent dosage.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a plot of turbidity (5 micron filtered NTU) and fluorescence
intensity
(5 micron filtered) vs. polymer dosage around the optimal polymer dosage for
Mississippi river water treated with poly(diallyldimethylammonium chloride) as
described in Example 4.
FIG. 2 is a plot of Particle Index data and fluorescence intensity vs. polymer
dosage for water treated with poly(diallyldimethylammonium chloride) at a
southern
U.S. petrochemical plant as described in Example 5. The data shows the optimal
polymer dosage to be about 0.5-0.7 ppm for this site.
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
3
FIG. 3 is a plot of the slope of the fluorescence curve in FIG. 2 vs. polymer
dosage. The data shows that the rate of fluorescence change is about 7 times
greater
around the optimal polymer dose than at over and under dose conditions.
FIG. 4 is a plot of the fluorescence change around the optimal traced polymer
dose at a southern U.S. petrochemical plant as described in Example 5 which
demonstrates how the dramatic fluorescence change around the optimal traced-
polymer dose can be used to automatically change the dose set point as the
influent
water quality fluctuates.
DETAILED DESCRIPTION OF THE INVENTION
This invention allows for treating agent residual monitoring by utilizing
fluorescent molecules. These molecules are selected such that they interact or
associate with the treating agent and any undesirable colloids. It is this
interaction that
partitions the fluorescent chromophore population between different
microenvironments. This partitioning changes the fluorescent properties such
that the
fluorescence intensity is notably different. The three microenvironments are
free
chromophore (i.e. dissolved in water), chromophore associated with treating
agent (i.e.
'bound' chromophore) and chromophore associated with the undesirable colloids.
When the treated water contains only freely dissolved chromophore the
fluorescence
intensity is what one would expect. When the treated water contains treating
agent
residuals and/or colloids, the fluorescence intensity is notably different
than expected
since the 'bound'/associated chromophores display different characteristics.
The
change in fluorescence intensity may be manifested by a non-proportional
increase or
decrease depending on the characteristics of the treated water, and the tagged
or traced
treating agent.
In a polymer overdose scenario filtrate water has more colloids and excess
polymer and excess tracer; together the tracer fluorescence intensity is
dramatically
different. In an under-dose scenario while colloids are present, no polymer or
tracer is
present; thus, the fluorescence intensity is low. Thus, in the transition
between treating
agent under-dose to over-dose scenario the fluorescence intensity notably
changes. It
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
4
is this change in fluorescence intensity that allows us to determine the
optimal treating
agent dose.
This difference between expected and actual fluorescence is used to
quantitatively estimate the treating agent residuals. By using fluorescence
response
and treating agent dose at two points, the treating agent residuals can be
estimated
from a calibration curve.
In an embodiment, the treating agent residuals can be estimated as a function
of
the difference between fluorescent molecules added to the water and
fluorescent
molecules detected in the water according to metric I.
LX1F2 _ X2F-1) I
(F2-Fi)
where xl and x2 are the first and second product dose and Fi and F2 are the
first and
second fluorescence measurements, in arbitrary units. In an embodiment, the
product
dose is in ppm.
In an embodiment, the treating agent residuals can be estimated as a function
of
the difference between the quenching expected and the quenching detected
according
to metric II.
~xl/F, -/F II
(1/F2 - 1/Fi)
where xl, x2, F1 and F2 are defined above. Fluorescence can be in arbitrary
units, but
can also be expressed as ppb of tracer molecule.
For purposes of this invention, overall quenching means any process or
processes that change the measured fluorescence such that the Stern-Volmer
plot as
described below is essentially linear. Put another way, "quenching" exists
when the
Stern-Volmer plot is linear.
In an embodiment, the residual concentration of the treating agent is
correlated
with treating agent dosage.
In an embodiment, the residual concentration of the treating agent is used to
determine an upper limit of treating dosage.
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
In an embodiment, the treating agent dosage is automatically maintained below
the upper limit,
In an embodiment optimal treating agent dosage is calculated using the
derivative of the inverse of fluorescence response with respect to treating
agent dose
5 and empirically correlating to any water quality parameter reflective of
system
perfonnance. Suitable water quality parameters include, but are not limited to
turbidity, silt density index (SDI), particle counts, and the like. In an
embodiment, this
correlation is accomplished using standard jar test methods to measure
fluorescence
and a water quality parameter such as turbidity and then calculating the
derivative of
the inverse of fluorescence response with respect to treating agent dose. Then
at the
acceptable water quality parameter dose point, the derivative of the inverse
fluorescence is the initial set point. Once implemented in full-scale, the set
point will
be fine-tuned for the optimal full-scale water process.
Inverse fluorescence can be related to quenching via Stern-Volmer plots. The
Stern-Volmer relationship is: 1/If= (1 +Kd[Q])/Io where: Kd = Quenching Rate
Constant, [Q] = Quencher(s) Concentration, Io = Fluorescence w/o Quenching, If
=
Measured Fluorescence. As noted above, for purposes of this invention,
quenching is
defined as occurring when the Stern-Volmer plot is essentially linear.
For purposes of this invention Io is assumed proportional to ppm of product
added(= k*ppm), then 1/If= (1/k + Kd[Q]/k)*(1/ppm) + 0. Here 'k' is a product
factor
that describes the concentration of the fluorescent molecule (i.e. the tracer)
in the
product being dosed (i.e. the treating agent). Therefore, if fluorescence
quenching is
occurring, a plot of 1/I fvs. 1/ppm is linear with a slope equal to (1/k +
Kd[Q])/k) and a
Y-intercept of zero.
In an embodiment, the set point of the derivative of the inverse fluorescence
response with respect to the treating agent dosage is used to automatically
control
treating agent dosage.
In an embodiment, an algorithm controls dosage iteration and calculates the
slope and residual function as defined above. Dosage iteration refers to a
method of
making a small adjustment in dosage, allowing the system to equilibrate, then
measuring some response.
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
6
More particularly, at a particular treating agent dosage (dosel), the treated
water's fluorescence (F1) is measured. The treating agent dosage is then
incremented
to slightly different dose (dose2) and the system is allowed to equilibrate.
At this new
dose, the fluorescence (F2) is measured. In an embodiment, the time required
for the
system to equilibrate is the retention time of the system, i.e. the time
needed for the
fluorescence to adjust to a change in treating agent dosage. Equilibration
time for
filtration systems is typically about five to about ten minutes but can be
longer
depending on the particular system.
At this point the slope of the inverse fluorescence vs. dosage curve is
calculated
with the algebraic relationship: slope =(1/F1- 1/F2)/(dosel - dose2). This
slope is
compared to a setpoint determined as described above and if it is greater than
the
setpoint, dosage is incrementally reduced, if it is less, the dose is
incrementally
increased. This is what we term slope control. Then the measured fluorescence
and
dose information is used to calculate the ideal slope and this is the initial
setpoint.
Once the full-scale system is activated, the setpoint is fine-tuned for
optimal system
performance. The foregoing is referred to as "manual set point determination.
Polyiner residuals are estimated by the fluorescence function using metrix I
or
II as described above and if the residuals are too high, the dosage is
automatically
reduced. In an embodiment, the residuals fluorescence function is used to
monitor the
system to ensure excessive residuals are not being fed to the treatment
system. In
another embodiment, the slope is used to automate the treating agent feed.
Thus, the
algorithm serves to maintain dosage control and insure that the treating agent
residuals
do not exceed an application specific set point.
For example, in reverse osmosis (RO) pretreatment systems, using iterative
control allows for dosage adjustment for changing influent waters, which is
different
from a fluorescence set point, which is valid only for a set influent. This
technology's
main advantage is the ability to monitor treating agent residuals and thus
allow for the
use of treating agents (a.k.a. - polyelectrolytes) for RO pretreatment to
reduce RO
influent's silt density index (SDI) and minimize cleaning cost, labor and lost
water
production.
Additionally, the fluorescence intensity of the process stream shows the
presence
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
7
of trace amounts of tagged or traced treating agent in the process stream.
This
fluorescence shows a sudden rapid change immediately following the optimal
dosage.
This rapid, non-proportional change in fluorescence can then be used as an
indication of
the optimal polymer dosage (the polymer dosage set point). See FIGS 2 and 4,
discussed
in detail below. As this set point can change in time due to, for example,
fluctuations in
the quality of the process water or operational upsets, the non-proportional
change in
fluorescence can be used to automatically change the set point, thereby
eliminating
manual intervention.
Accordingly, in another embodiment, the change in the measured fluorescence
intensity of the process water at the first and second treating agent doses is
correlated
with a non-proportional change in measured fluorescence, wherein the non-
proportional change in measured fluorescence of the process water is used to
determine a set point corresponding to the optimal treating agent dosage.
According to this embodiment, the set point can be automatically determined
by starting the system with a suboptimal dose (may be 0 ppm) of tagged or
traced
treating agent and measuring and data logging the initial mean value of
fluorescence,
F1, and the corresponding polymer dose, P1. The polymer dose is then
incrementally
increased, for example by about 0.1 ppm to a value P2 and the system is
allowed to
equilibrate as described above. The new mean fluorescence value (172) and the
corresponding traced polymer dose, P2 are data logged. In connection with the
fluorescence values the operator should decide on the confidence level the
data has to
satisfy to be recognized as genuine change (for example, 95 percent confidence
that a
change in fluorescence represents a real change based on the system
parameters).
The treating agent dosage P2 is then incrementally increased until a point of
non-proportional change in fluorescence (inflection point) corresponding to
the
optimal treating agent dose occurs. For example, such a point of non-
proportional
change could occur when F2 is more than about 50 percent of F1 as large
changes in
fluorescence for an incremental increase in traced or tagged treating agent
are not
expected outside the optimal dosage region.
After a suitable amount of time, for example 30 minutes, data log the new
mean fluorescence value (F3) and the corresponding traced polymer dose (P3).
If F3 is
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
8
greater than F2 by a significant amount the polymer dose should be decreased.
If F3 is
significantly smaller than F2 the polymer dose should be increased to return
to the
inflection point.
Suitable data logging protocols should measure the variable of interest (eg.,
fluorescence) and accumulate 100-300 data points; calculate the mean and the
standard
deviation for the measurement; ensure that the standard deviation is low (for
example
less than about 10 percent of the mean); and store the measured mean (F;), the
standard
deviation (S;) and the number of data points (N;) used to calculate the above.
In an embodiment, the set point is automatically changed to account for
fluctuations in the process water quality.
In an embodiment, the treating agent dosage is automatically maintained at the
set point.
In an embodiment, the treating agent is traced with one or more fluorescent
tracers. These fluorescent tracers may or may not be appreciably or
significantly
affected by any other chemistry in the water treatment process, or by the
other system
parameters such as pH, temperature, ionic strength, redox potential,
microbiological
activity or biocide concentration. As long as the chemistry in the water
treatinent
process does not significantly change during the retention time (usually about
ten
minutes), the control algorithm automatically accounts for significant changes
in
fluorescence.
The fluorescent tracers must be transportable with the water treatment process
water and thus are substantially, if not wholly, water-soluble therein at the
use
concentration, under the temperature and pressure conditions specific and
unique to the
water treatment process.
Representative fluorescent tracers include, but are not limited to tracers
described in U.S. Patent No. 6,730,227, incorporated herein by reference.
In an embodiment, the tracers are selected from fluorescein, sodium salt (CAS
Registry No. 518-47-8, aka Acid Yellow 73, Uranine); 1,5-naphthalenedisulfonic
acid
disodium salt (hydrate) (CAS Registry No. 1655-29-4, aka 1,5 - NDSA hydrate);
xanthylium, 9-(2,4-dicarboxyphenyl)-3,6-bis(diethylamino)-, chloride, disodium
salt,
also known as Rhodamine WT (CAS Registry No. 37299-86-8); 1-deoxy-1-(3,4-
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
9
dihydro-7,8-dimethyl-2,4-dioxobenzo[g]pteridin-10(2H)-yl)-D- ribitol, also
known as
Riboflavin or Vitamin B2 (CAS Registry No. 83-88-5); fluorescein (CAS Registry
No.
2321-07-5); 2-anthracenesulfonic acid sodium salt (CAS Registry No. 16106-40-
4);
1,5 -anthracenedisulfonic acid (CAS Registry No. 61736-91-2) and salts
thereof; 2,6-
anthracenedisulfonic acid (CAS Registry No. 61736-95-6) and salts thereof; 1,8-
anthracenedisulfonic acid (CAS Registry No. 61736-92-3) and salts thereof; and
mixtures thereof. The fluorescent tracers listed above are commercially
available from
a variety of different chemical supply companies.
In an embodiment, the treating agent is tagged with a fluorescent moiety, for
example by incorporating the fluorescent moiety into a polymeric treatment
polymer
itself, or by post modification of a treatment polymer with a fluorescent
moiety
capable of forming a covalent bond with the treatment polymer. The preparation
and
use of polymers containing a fluorescent moiety is described in, for exainple,
U.S.
Patent Nos. 6,312,644; 6,077,461; 5,986,030; 5,998,632; 5,808,103; 5,772,894;
5,958,788 and PCT US01/81654, incorporated herein by reference.
The dosage of the fluorescent tracer is an amount that is at least sufficient
to
provide a measurable concentration in the treated water. Typical doses range
from
about 50 ppt (parts per trillion) to about 100 ppb (parts per billion),
preferably from
about 0.1 ppb to about 10 ppb, based on fluorescent agent concentration. Note
that 50
ppt is about the detection limit of currently available industrial
fluorometers.
Improvements in fluorometer technology are likely to reduce this detection
limit and
are envisioned.
The fluorescent tracers can be detected by utilizing a variety of different
and
suitable techniques. For example, fluorescence emission spectroscopy on a
substantially continuous basis, at least over a given time period, is one of
the preferred
analytical techniques according to an embodiment of this invention. One method
for
the continuous on-stream measuring of chemical tracers by fluorescence
emission
spectroscopy and other analysis methods is described in U.S. Patent No.
4,992,380,
incorporated herein by reference.
Examples of fluorometers that may be used in the practice of this invention
include the Xe II and TRASAR 8000 fluorometer (available from Nalco Company,
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
Naperville, IL); the Hitachi F-4500 fluorometer (available from Hitachi
through
Hitachi Instruments Inc., San Jose, CA); the JOBIN YVON FluoroMax-3 "SPEX"
fluorometer (available from JOBIN YVON Inc., Edison, NJ); and the Gilford
Fluoro-
IV spectrophotometer or the SFM 25 (available from Bio-tech Kontron through
5 Research Instruments International, San Diego, CA). It should be appreciated
that the
foregoing list is not comprehensive and is intended only to show examples of
representative fluorometers. Other commercially available fluorometers and
modifications thereof can also be used in this invention.
It should be appreciated that a variety of other suitable analytical
techniques
10 may be utilized to measure the amount of fluorescent tracers. Examples of
such
techniques include combined HPLC-fluorescence analysis, colorimetry analysis,
ion
selective electrode analysis, transition metal analysis, chemiluminescence,
pulsed
fluorescence measurements, and the like.
In an embodiment, the present invention includes a controller programmed with
the foregoing algorithm and which continuously (i.e. within the tiinescale of
the
retention time, typically every few minutes) makes incremental changes in the
treating
agent dosage and performs the calculations described above so as to maintain
the
treating agent residuals at the desired set point.
The controller can be configured and/or adjusted in a variety of different and
suitable ways. Alternative methods could include using three or more points to
measure the fluorescence response and then use analytical curve fitting
methods to
determine optimal dosage.
The controller can be either hard wired (e.g., electrical communication
cable),
or can communicate with the other components described herein by wireless
communication (e.g., wireless RF interface), a pneumatic interface and the
like.
As described above, this invention is a method of monitoring treating agent
residuals and controlling treating agent dosage in water treatment processes.
"Treating
agent" is meant herein without limitation to include treatment chemicals that
enhance
solid-liquid separation, membrane separation process performance, antiscalants
that
retard/prevent scale formation and deposition on surfaces in contact with the
treated
water, antifoulants that retard/prevent membrane fouling, biodispersants,
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
11
microbial-growth inhibiting agents such as biocides and cleaning chemicals
that
remove deposits from surfaces that contact the treated water.
The present invention is applicable to all industries that can employ water
treatment processes. For example, the different types of industrial processes
in which
the method of the present invention can be applied generally include raw water
processes, waste water processes, industrial water processes, municipal water
treatment, food and beverage processes, pharmaceutical processes, electronic
manufacturing, utility operations, pulp and paper processes, mining and
mineral
processes, transportation-related processes, textile processes, plating and
metal
working processes, laundry and cleaning processes, leather and tanning
processes, and
paint processes.
In particular, food and beverage processes can include, for example, dairy
processes relating to the production of cream, low-fat milk, cheese, specialty
milk
products, protein isolates, lactose manufacture, whey, casein, fat separation,
and brine
recovery from salting cheese. Uses relating to the beverage industry
including, for
example, fruit juice clarification, concentration or deacidification,
alcoholic beverage
clarification, alcohol removal for low-alcohol content beverages, process
water; and
uses relating to sugar refining, vegetable protein processing, vegetable oil
production/processing, wet milling of grain, animal processing (e.g., red
meat, eggs,
gelatin, fish and poultry), reclamation of wash waters, food processing waste
and the
like.
Examples of industrial water uses as applied to the present invention include,
for example, boiler water production, process water purification and
recycle/reuse,
softening of raw water, treatment of cooling water blow-down, reclamation of
water
from papermaking processes, desalination of sea and brackish water for
industrial and
municipal use, drinking/raw/surface water purification including, for example,
the use
of membranes to exclude harmful micro-organisms from drinking water, polishing
of
softened water, membrane bio-reactors, mining and mineral process waters.
Examples of waste water treatment applications with respect to the method of
this invention include, for exainple, industrial waste water treatment,
biological waste
treatment systems, removal of heavy metal contaminants, polishing of tertiary
effluent
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
12
water, oily waste waters, transportation related processes (e.g., tank car
wash water),
textile waste (e.g., dye, adhesives, size, oils for wool scouring, fabric
finishing oils),
plating and metal working waste, laundries, printing, leather and tanning,
pulp and
paper (e.g., color removal, concentration of dilute spent sulfite liquor,
lignin recovery,
recovery of paper coatings), chemicals (e.g., emulsions, latex, pigments,
paints,
chemical reaction by-products), and municipal waste water treatment (e.g.,
sewage,
industrial waste).
Other examples of industrial applications of the present invention include,
for
example, semiconductor rinse water processes, production of water for
injection,
pharmaceutical water including water used in enzyme production/recovery and
product
fonnulation, and electro-coat paint processing.
In an embodiment, the present invention is applied in raw or treated water
applications where the filtrate is used as feed for reverse osmosis units. It
is
particularly important that polymer residuals not foul RO membranes, although
the
present invention is envisioned for any application where the use pretreatment
polymers is desired, but excessive polymer residuals are not, as well as
applications
that benefit from control of treating agent dosing. Some examples would be,
surface
water clarification, ground water clarification, tertiary treatment of
wastewater, and
seawater clarification. The product of such clarification processes could be
used for,
but not limited to, industrial process water, boiler or cooling water make-up
water, or
residential water.
In an embodiment, the water treatment process is a solid-liquid separation
process.
In an embodiment, the solid-liquid separation process comprises treatment of
the water with one or more coagulants or flocculants, or a combination
thereof, to form
a mixture of water and coagulated and flocculated solids and separation of the
coagulated and flocculated solids from the water.
Suitable flocculatants include high molecular weight cationic, anionic,
nonionic, zwitterionic or amphoteric polymers. Suitable flocculants generally
have
molecular weights in excess of 1,000,000 and often in excess of 5,000,000. The
polymeric flocculant is typically prepared by vinyl addition polymerization of
one or
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
13
more cationic, anionic or nonionic monomers, by copolymerization of one or
more
cationic monomers with one or more nonionic monomers, by copolymerization of
one
or more anionic monomers with one or more nonionic monomers, by
copolymerization
of one or more cationic monomers with one or more anionic monomers and
optionally
one or more nonionic monomers to produce an amphoteric polymer or by
polymerization of one or more zwitterionic monomers and optionally one or more
nonionic monomers to form a zwitterionic polymer. One or more zwitterionic
monomers and optionally one or more nonionic monomers may also be
copolymerized
with one or more anionic or cationic monomers to impart cationic or anionic
charge to
the zwitterionic polymer.
While cationic polymer flocculants may be formed using cationic monomers, it
is also possible to react certain non-ionic vinyl addition polymers to produce
cationically charged polymers. Polymers of this type include those prepared
through
the reaction of polyacrylamide with dimethylamine and formaldehyde to produce
a
Mannich derivative.
Similarly, while anionic polymer flocculants may be formed using anionic
monomers, it is also possible to modify certain nonionic vinyl addition
polymers to
form anionically charged polymers. Polymers of this type include, for example,
those
prepared by the hydrolysis of polyacrylamide.
The flocculants may be used in the solid form, as an aqueous solution, as a
water-in-oil emulsion, or as dispersion in water. Representative cationic
polymers
include copolymers and terpolymers of (meth)acrylamide with dimethylaminoethyl
methacrylate (DMAEM), dimethylaminoethyl acrylate (DMAEA), diethylaminoethyl
acrylate (DEAEA), diethylaminoethyl methacrylate (DEAEM) or their quaternary
ammonium forms made with dimethyl sulfate, methyl chloride or benzyl chloride.
Water-soluble coagulants are well known and commercially available. Suitable
coagulants may be inorganic or organic. Representative inorganic coagulants
include
alum, sodium aluminate, polyaluminum chlorides or PACIs(which also may be also
be
referred to as aluminum chlorohydroxide, alurninum hydroxide chloride, basic
aluminum chloride and polyaluminum hydroxychloride, and the like), sulfated
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
14
polyaluminum chlorides, polyaluminum silica sulfate, ferric sulfate, ferric
chloride,
and the like and blends thereof.
Many water-soluble organic coagulants are formed by condensation
polymerization. Examples of polymers of this type include epichlorohydrin-
dimethylamine, and epichlorohydrin-dimethylamine-ammonia polymers.
Additional coagulants include polymers of ethylene dichloride and ammonia,
or ethylene dichloride and dimethylamine, with or without the addition of
ammonia,
condensation polymers of multifunctional amines such as diethylenetriamine,
tetraethylenepentamine, hexamethylenediamine and the like with
ethylenedichloride
and polymers made by condensation reactions such as melamine formaldehyde
resins.
Additional coagulants include cationically charged vinyl addition polymers
such as polymers and copolymers of diallyldimethylammonium chloride,
dimethylaminoethyhnethacrylate, dimethylaminoethylmethacrylate methyl chloride
quaternary salt, methacrylamidopropyltrimethylammonium chloride,
(methacryloxyloxyethyl)trimethyl ammonium chloride, diallylmethyl(beta-
propionamido)ammonium chloride, (beta-methacryloxyloxyethyl)trimethyl-
aminonium methylsulfate, quaternized polyvinyllactam, dimethylamino-
ethylacrylate
and its quaternary ammonium salts, vinylamine and acrylamide or methacrylamide
which has been reacted to produce the Mannich or quaternary Mannich
derivatives.
The molecular weights of these cationic polyiners, both vinyl addition and
condensation, range from as low as several hundred to as high as one million.
Preferably, the molecular weight range should be from about 20,000 to about
1,000,000.
The selection of the proper flocculant and coagulant for a particular
application
and determination of the effective dose may be empirically determined by one
of skill
in the art of water treatment based on the characteristics of the particular
water being
treated.
The coagulated and flocculated solids may then be separated from the water by
any of a number of means available in the art of solid-liquid separation
including
clarifiers, by centrifuges, dissolved air flotation, mechanical means such as
belt press
or plate and frame press, and membrane filtration or media filtration.
Membrane
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
filtration is generally considered micro or ultra filtration involving pliable
membranes,
ceramic membranes and the like. Media filtration is generally considered to be
any
granular media involved as a barrier to contaminants in water and they are
commonly
sand, anthracite, and garnet. Any media that functions as a barrier is
envisioned and
5 this includes, but is not limited to, micro and macro breads, powders,
activated carbon,
ceramics, etc.
In an embodiment, the solid-liquid separation process is a membrane separation
process wherein the coagulated and flocculated solids are separated from the
water by
filtration through a membrane Residuals monitoring and dosage control
according to
10 this invention can be used to enhance the operational efficiency of
membrane filtration
systems as described below.
Membrane separations commonly used for water purification or other liquid
processing include microfiltration (MF), ultrafiltration (UF), nanofiltration
(NF),
reverse osmosis (RO), electrodialysis, electrodeionization, pervaporation,
membrane
15 extraction, membrane distillation, membrane stripping, membrane aeration,
and other
processes. The driving force of the separation depends on the type of the
membrane
separation. Pressure-driven membrane filtration, also known as membrane
filtration,
includes microfiltration, ultrafiltration, nanofiltration and reverse osmosis,
and uses
pressure as a driving force, whereas the electrical driving force is used in
electrodialysis and electrodeionization.
In an embodiment, the membrane separation process comprises one or more
pretreatment steps wherein a portion of the coagulated and flocculated solids
are
separated from the water prior to filtration of the water through a
nanofiltration and/or
reverse osmosis membrane.
In an embodiment, the membrane separation system is a reverse osmosis
system.
In reverse osmosis, the feed stream is typically processed under cross flow
conditions. In this regard, the feed stream flows substantially parallel to
the membrane
surface such that only a portion of the feed stream diffuses through the
membrane as
permeate. The cross flow rate is routinely high in order to provide a scouring
action
that lessens membrane surface fouling. This can also decrease concentration
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
16
polarization effects (e.g., concentration of solutes in the reduced-turbulence
boundary
layer at the membrane surface, which can increase the osmotic pressure at the
membrane and thus can reduce permeate flow). The concentration polarization
effects
can inhibit the feed stream water from passing through the membrane as
permeate,
thus decreasing the recovery ratio, e.g., the ratio of permeate to applied
feed stream. A
recycle loop(s) may be employed to maintain a high flow rate across the
membrane
surface.
Reverse osmosis processes can employ a variety of different types of
membranes. Such commercial membrane element types include, without limitation,
hollow fiber membrane elements, tubular membrane elements, spiral-wound
membrane elements, plate and frame membrane elements, and the like, some of
which
are described in more detail in "The Nalco Water Handbook," Second Edition,
Frank
N. Kemmer ed., McGraw-Hill Book Company, New York, N.Y., 1988, incorporated
hereinto, particularly Chapter 15 entitled "Membrane Separation". It should be
appreciated that a single membrane element may be used in a given membrane
filtration system, but a number of membrane elements can also be used
depending on
the industrial application.
A typical reverse osmosis system is described as an example of membrane
filtration and more generally membrane separation. Reverse osmosis uses mainly
spiral wound elements or modules, which are constructed by winding layers of
semi-
porous membranes with feed spacers and permeate water carriers around a
central
perforated permeate collection tube. Typically, the modules are sealed with
tape
and/or fiberglass over-wrap. The resulting construction has one channel, which
can
receive an inlet flow. The inlet stream flows longitudinally along the
membrane
module and exits the other end as a concentrate stream. Within the module,
water
passes through the semi-porous membrane and is trapped in a permeate channel
which
flows to a central collection tube. From this tube it flows out of a
designated channel
and is collected.
In practice, membrane modules are stacked together, end-to-end, with inter-
connectors joining the permeate tubes of the first module to the permeate tube
of the
second module, and so on. These membrane module stacks are housed in pressure
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
17
vessels. Within the pressure vessel feed water passes into the first module in
the stack,
which removes a portion of the water as permeate water. The concentrate stream
from
the first membrane becomes the feed stream of the second membrane and so on
down
the stack. The permeate streanis from all of the membranes in the stack are
collected
in the joined permeate tubes.
Within most reverse osmosis systems, pressure vessels are arranged in either
"stages" or "passes." In a staged membrane system, the combined concentrate
streams
from a bank of pressure vessels are directed to a second bank of pressure
vessels where
they become the feed stream for the second stage. Commonly systems have 2 to 3
stages with successively fewer pressure vessels in each stage. For example, a
system
may contain 4 pressure vessels in a first stage, the concentrate streams of
which feed 2
pressure vessels in a second stage, the concentrate streams of which in turn
feed 1
pressure vessel in the third stage. This is designated as a"4:2:1" array. In a
staged
membrane configuration, the combined permeate streams from all pressure
vessels in
all stages are collected and used without further membrane treatment. Multi-
stage
systems are used when large volumes of purified water are required. The
permeate
streams from the membrane system may be further purified by ion exchange or
other
means.
In a multi-pass system, the permeate streams from each bank of pressure
vessels are collected and used as the feed to the subsequent banks of pressure
vessels.
The concentrate streams from all pressure vessels are combined without further
membrane treatment of each individual stream. Multi-pass systems are used when
very high purity water is required, for example in the microelectronics or
pharmaceutical industries.
It is well known to those skilled in the art, that various coagulants are
needed to
maximize the efficiency of solid-liquid separation. As noted above, amongst
these
suitable coagulants are aluminum and iron compounds and synthetic
polyelectrolytes.
Unfortunately, using aluminum, such as alum, produces residuals that form
intractable
scale on membranes. Polyelectrolytes are frequently used for general
clarification, but
since RO membranes are anionic polyamide films and polyelectrolyte coagulants
are
cationic, it is widely feared that polymer will deposit on membranes via
electrostatic
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
18
attraction and cause permanent fouling. This situation would require
expensive,
inefficient membrane replacement. Therefore, in reverse osmosis filtration
systems,
pretreatment is critical to efficient operation.
Reverse osmosis pretreatment schemes will vary with the type of water. For
example, waters that have greater than ca. 10 NTU will usually utilize
sedimentation
followed by filtration. Waters that are cleaner than ca. 10 NTU can use direct
filtration
techniques.
Filtration generally consists of a media filter or a membrane micro or
ultrafilter.
Media filters consist of particulate solids on the order of 1 mm diameter.
While a wide
variety of materials can be used, the most common materials are sand, garnet
and
anthracite singly or in combination. Micro and ultrafilters can consist of
either
ceramic or membrane construction and have significantly smaller pore size
compared
to media filters. All three of these types are used for RO pretreatment.
Various coagulants can be used as filter aids in RO pretreatment. Filter aids
function by modifying the influent's particle size and surface properties in
order to
facilitate particulate capture by the filter. Which type of filter aid to use
varies from
water to water and for optimal filtration, the proper chemistry, or even mix
of
chemistries, is critical.
In an embodiment, this invention is a reverse osmosis pretreatment system
wherein the portion of coagulated and flocculated solids is removed from the
water by
filtration through a media filter.
In an embodiment, this invention is a reverse osmosis pretreatment system
wherein the filter aid is one or more coagulants selected from alum,
polyaluminum
chloride, ferric chloride, ferric sulfate, poly(diallyldimethylammonium
chloride) and
Epi-DMA.
In an embodiment, this invention is a reverse osmosis pretreatment system
wherein the fluorescent tracer is selected from fluorescein, rhodamine B,
rhodamine
WT and 1,3,6,8-pyrenetetrasulfonic acid tetrasodium salt.
In an embodiment, this invention is a reverse osmosis pretreatment system
wherein the fluorescent tracer is poly(diallyldimethylammonium chloride)
tagged with
luminol, rhodamine or fluorescein.
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
19
The foregoing may be better understood by reference to the following
examples, which are presented for purposes of illustration and are not
intended to limit
the scope of the invention.
Example 1
As discussed above, the optimal treating agent dosage is calculated using the
inverse derivative of fluorescence response with respect to treating agent
dose and
correlating to a water quality parameter, in this case turbidity, using
standard jar test
methods to measure fluorescence and turbidity.
For purposes of this example, jar tests are accomplished using a four-unit jar
tester from A&F Machine Products Co., Berea, OH (model number "JAR MIXER")
according to the following protocol.
1) Place a 250-1,000 mL test sample in a sample jar and initiate stirring at
200
rpm.
2) Add treating agent via syringe into the vortex of the stirred sample and
continue stirring for 30 seconds.
3) Slow the stirring to 15 to 60 rpm and continue stirring for 5 minutes.
4) Stop stirring, remove the paddles from the sample and allow the sample to
settle for 5 minutes.
5) Remove a sample of the supernatant via pipette or syringe from a level
about 1
cm below the sample surface and filter the sample through a 5 m syringe
filter
directly into the turbidity and fluorescence sample cells for measurement of
the
appropriate property.
The treating agent promotes the agglomeration of smaller particles into larger
particles, or flocs, that can be more readily separated from the water, for
example by
settling or filtration. In the event that visible flocs are formed, the floc
size of the
average particle formed may be ranked versus the benchmark. Other factors such
as
water clarity between floc particles, floc shape, tightness, etc. may also be
noted for
comparison.
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
All jar tests should be benchmarked using a plant's current coagulation
program. This allows for adjustment of testing parameters to correlate with
full-scale
performance. The actual dose in use in the plant at the time of the tests is
set as the
dose benchmark for the series of tests and the clarifier overflow turbidity
(or color,
5 etc.) is the performance benchmark.
The parameters of the jar test can be varied based on the treatment
application
and characteristics of the samples being tested. Accordingly, alternative
methods may
employ a longer fast mix of 1 to 5 minutes and a slow mix of 2-10 minutes with
or
without a settling stage. For example, if alum is added at the intake and
travels a long
10 distance, then the fast mix will be longer than if alum is added only 20
feet before the
clarifier. A plant that is hydrologically overloaded will need a slow mix
shorter slow
mix. These factors are varied based on experience of treatment personnel.
For this example 1, Mississippi River water is jar tested using 30 ppm of a 50
weight percent ferric sulfate solution and augmenting the ferric sulfate
coagulant with
15 a fluorescein-traced poly(diallyldimethylammonium chloride) ("polyDADMAC")
coagulant. The polyDADMAC is approximately 20-weight percent polymer and 0.19
weight percent fluorescein with the balance being an approximately one weight
percent
saline solution. The polyDA.DMAC solution concentration is listed in Table 1
in ppm
on a weight/weight ("w/w") basis. The performance metrics are turbidity in NTU
and
20 detected fluorescein fluorescence in ppb (w/w) of fluorescein in the
filtrate. The
Fluorometer is a Hitachi Model F-4500 and calibration is accomplished using
fluorescein in deionized water. A Hach portable turbidimeter Model 2100P is
used for
NTU measurements. The results are summarized in Table 1.
Table 1
Settled Fluorescence
ppm 1/ m NTU Intensity, bQuenchin
1 1.00000 10.9 2.08 0.48077
5 0.20000 4.3 8.65 0.11561
9 0.11111 5.0 15.3 0.06536
13 0.07692 6.9 21.9 0.04566
17 0.05882 11.0 27.3 0.03663
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
21
In Table 1, the last column is quenching which is analytically calculated as
the
inverse of the fluorescence. Therefore the quenching has units of ppb"1. Note
that the
fluorescence units can be arbitrary and for the purposes of the invention,
only relative
fluorescence intensity is required.
Example 2
Testing is also accomplished in the field according to the method of Example 1
using the Hach 2100P turbidimeter on Mississippi River water that has been
previously
treated for turbidity removal resulting in a turbidity of about 1 NTU.
Fluorescence
data is obtained using a TRASAR 8000 Fluorometer from Nalco Company,
Naperville,
IL. The TRASAR 8000 requires a correction at zero added fluorescein with the
correction being a subtraction of 0.06 ppb fluorescein. This correction is
attributed to
less precise optics versus the research grade Hitachi. Po1yDADMAC is the
treating
agent. The fast mix is 30 seconds, slow mix three minutes and settling time
five
minutes. The Fluorescence Intensity, ppb is without the 0.06 ppb fluorescein
background correction, while the Corrected Quenching is the inverse of the
corrected
Fluorescence Intensity, so it has units of ppb-1. The results are shown in
Table 2
Table 2
5um
Filtered FluorescenCorrected
m 1/ppm NTU Intensit , bQuenchin
0.4 2.5000 0.30 0.1140 18.5185
0.6 1.6667 0.43 0.1470 11.4943
0.8 1.2500 0.69 0.1650 9.5238
1.0 1.0000 0.50 0.3300 3.7037
Note that in Table 2, 0.06 ppb is subtracted to correct the fluorescence.
Example 3
Natural surface water from a lake in Montana, USA, is treated with a dual
filter
aid program using ferric sulfate solution and polyDADMAC, as summarized in
Table
3. The fast mix is two minutes, slow mix ten minutes and no settling time and
the
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
22
0.057 ppb fluorescein correction is subtracted from the measured fluorescence
intensity. The results are summarized in Table 3.
Table 3
5um
Filtered Fluorescence Corrected
ppm 1/ppm NTU Intensity, bQuenchin
0.50 2.0000 1.05 0.408 2.8490
1.00 1. 0000 0.74 0.615 1.7921
1.50 0.6667 0.64 0.792 1.3605
2.00 0.5000 0.52 1.260 0.8313
Note that in Table 3, 0.057 ppb is subtracted to correct the fluorescence.
As shown in Table 4, non-weighted regression analysis of Stern-Volmer plots
for the above data have 'goodness-of-fit' (r2) of approximately 90% for the
field data
and 99% for the laboratory data. Data using the Hitachi research grade
Fluoroineter
presumably has a higher fit due to more accurate fluorescence measurements
with
significantly lower background light being detected.
Table 4
Example #1 Example #2 Example #3
99% 89% 89%
Kd[Q] 91.5 642.4 141.3
These data suggest that the quenching rate constant multiplied by the quencher
concentration varies with the water being treated. This variability with
respect to the
water tested suggests that quenching may be related to water treatability such
that
quenching can be advantageously used as a water treatment parameter.
Example 4
For this Example, Mississippi River water is jar tested using
poly(diallyldimethylainmonium chloride) ("polyDADMAC") coagulant according to
the method of Example 1. The polyDADMAC is approximately 20-weight percent
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
23
polymer with the balance being an approximately one weight percent saline
solution.
The polyDADMAC solution concentration is listed in Table 5 in ppm on a
weight/weight ("w/w") basis. The performance metrics are turbidity in NTU and
detected fluorescein fluorescence in ppb (w/w) of fluorescein in the filtrate.
The
Fluorometer is a Hitachi Model F-4500 and calibration is accomplished using
fluorescein in deionized water. A Hach portable turbidimeter Mode12100P is
used for
NTU measurements. The results are summarized in Table 5.
Table 5
Direct Filtration Jar Testin
F'otymer dose, 0.5 1.0 1.5
ppm
5mm Filtered NTU 0.70 0.62 0.66
5mm Filtered 0.101 0.179 0.833
fluorescein, b
The optimum polymer dosage, using these low-resolution jar tests, is found to
be about 1.0 ppm of polyDADMAC (clearest water with the lowest turbidity,
Table 5).
As polymer dose is increased over the optimum (lppm in this particular case)
the
turbidity of the filtrate increases again, i.e., the filtrate water quality
diminishes (Table
5). Simultaneously, the fluorescence intensity of the filtered water increases
disproportionately rapidly (See FIG. 1). This occurs because the overdosed
polymer
breaks through to the filtrate side enhancing the fluorescence intensity. The
more
polymer in the filtrate the greater is the fluorescence enhancement. Small
amounts of
polymer (e.g., 0.5 ppm of polyDADMAC as a 20% active product) significantly
increases fluorescein fluorescence intensity (Fig. 1). This enhanced
fluorescence can
be used to indicate the presence of excess residual polymers.
Example 5
A field trial is conducted in a southern U.S. petrochemical plant. The plant
takes in Mississippi River water and uses lime softening for clarification.
Water is
then pH adjusted to 7 and chlorinated with bleach. Next, the feed water is
heated to 78
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
24
F in a plate heat exchanger. Feed residual oxidant of 0.4-0.6 ppm is typically
observed down stream from the heat exchanger.
Traced polymer is added as a filter-aid ahead of the multi-media filter. It is
prepared by blending, 99.77% PoIyDADMAC and 0.23% fluorescein by weight.
Water from the heat exchanger is treated with traced po1yDADMAC and pumped
through a static mixer and into three down-flow multimedia filters. Polymeric
filter
aid is fed using a 24-gpd Series E+ Pulsafeeder (model LPB4MA-VTCl-XXX) pump
with 4-20 mA control. After a 25-minute delay, the filtrate fluorescence is
measured
using a Trasar 340 fluorometer (Nalco Company part # TSR341) and a VersaTrak
PLC
controller (Sixnet, Clifton Park, NY, part # VT-A3-422-44P). The residence
time of
the multi media filters is about 25 minutes, which is the time a very small
change talces
to propagate through the system.
Chemtrac Systems Inc.'s "Particle Monitor" (Nalco Company part# 041-
PM25011.88) measures the "particle Index" as an indicator of the
particulate/contaminants content in the filtrate. This system can be used to
monitor the
quality of low solids, low turbidity systems. The higher the Particle Index
the higher
the particulate/contaminants content in the filtrate.
The Detection accuracy is as follows. Particle monitor: The standard deviation
of the reading is about 15% of the mean. Changing the polymer dose the trend
in
Particle Index data is very clear. Field trial: The standard deviation of the
reading is
about 10% of the mean. The accuracy can be further improved by increasing the
concentration of fluorescein in the product.
The optimal polyDADMAC polymer dosage to the multimedia filter influent is
between 0.5-0.7ppm. In this range the particle monitor has the lowest reading
(See
FIG. 2), signifying that a minimum amount of particles/contaminants is present
in the
filtrate. Below this polymer dosage more colloids escape with the filtrate and
above
this dosage polymer breakthrough occurs, both of which increase the particle
monitor
reading. Similar to the observation in Table 1, FIG. 2 shows that just beyond
the
optimal dosage, a non-proportional increase in the fluorescence intensity of
the filtrate
occurs. This increase is due to the excess traced polymer in the filtrate,
just as was
CA 02595337 2007-07-19
WO 2006/078847 PCT/US2006/001948
observed in the jar test results. This sharp increase in fluorescence is
statistically
>97% significant and is a strong indicator of optimal traced-polymer dosage
level.
FIG. 3 is a plot of the slope of the fluorescence curve in FIG. 2. Around the
optimum traced-polymer dose the fluorescence slope is 7 times higher than the
5 surrounding under and over dose conditions. This makes it particularly
attractive to be
used as a marker to determine optimal polymer dosage.
Thus, as the influent water quality changes the optimum polymer dose set point
can be changed automatically (See FIG. 4). When dosing a traced-polymer such
that
filtrate fluorescence is at its inflection point, changes in water quality
result in dramatic
10 changes in fluorescence, which are easily measurable and usable to adjust
the set point.
This method eliminates the need for manual interference to automatically track
and
dose the optimum amount of traced-polymer.
It should be understood that various changes and modifications to the
presently
preferred embodiments described herein will be apparent to those skilled in
the art.
15 Such changes and modifications can be made without departing from the
spirit and
scope of the present invention and without diminishing its attendant
advantages. It is
therefore intended that such changes and modifications be covered by the
appended
claims.
