i USE OF REVERSE OSMOSIS TO RECOVER WATER FROM A NUTRIENT SEPARATION SYSTEM FOR DAIRY MANURE MANAGEMENT By John Stephen Budaj A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Biosystems Engineering - Master of Science 2016
105
Embed
USE OF REVERSE OSMOSIS TO RECOVER WATER …3961/datastream/OBJ/...USE OF REVERSE OSMOSIS TO RECOVER WATER FROM A NUTRIENT SEPARATION SYSTEM FOR DAIRY MANURE ... Example of a cleaning
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
i
USE OF REVERSE OSMOSIS TO RECOVER WATER FROM A NUTRIENT SEPARATION SYSTEM FOR
DAIRY MANURE MANAGEMENT
By
John Stephen Budaj
A THESIS
Submitted to Michigan State University
in partial fulfillment of the requirements for the degree of
Biosystems Engineering - Master of Science
2016
ii
ABSTRACT
USE OF REVERSE OSMOSIS TO RECOVER WATER FROM A NUTRIENT SEPARATION SYSTEM FOR DAIRY MANURE MANAGEMENT
By
John Stephen Budaj
Manure storage is sometimes limited and over application of manure on fields is dangerous
to the environment, especially ground water. Filtration systems using reverse osmosis (RO), are one
alternative approach to reusing waste water and eliminating risk for manure spills or harming the
environment. There are various challenges involved when trying to optimize this system, one of
which is membrane fouling upon the RO filters. The rapid increase in pressure and reduced
permeate generation due to fouling is problematic. To address this issue, antiscalant was added to
the feed and the feed stream was pH adjusted using sulfuric acid or hydrochloric acid. In order to
determine what foulants were present on the membrane, some of the membranes were dissected
and tested using scanning electron microscopy (SEM), energy dispersive x-ray (EDX), chromatic
indicated organic and silcate scaling were present under all operating conditions. In addition, there
was a greater degree of scaling using non pH adjusted feed versus pH adjusted. The use of air
stripped water processed through the RO system was used for all the experiments at the pH of 5.5,
6.5, 7.5 and 8. Sulfuric acid and hydrochloric acid were used to lower the pH of the air stripped
water (normal pH of 8) and the various acid runs were compared in the study. After optimizing the
system, and using permeate production as a gauge, the system performed best at a feed pH of 6.5
using sulfuric acid and an appropriate dose of antiscalant, determined by the Avista Advisor
modeling software. The use of hydrochloric acid was very expensive when pH adjusting versus the
use of sulfuric acid for pH adjustment. Operational costs and capital costs were also determined.
iii
ACKNOWEDGEMENTS
I would like to acknowledge and thank my major professor Dr. Saffron, my mentor Jim
Wallace, my committee member Dr. Safferman, and the great staff in the Biosystems department at
MSU for all their input and assistance during my project.
iv
TABLE OF CONTENTS
LIST OF TABLES ...................................................................................................................................... vi
LIST OF FIGURES ................................................................................................................................... vii
KEY TO ABBREVIATIONS ......................................................................................................................... x
3.1 Initial Observations ......................................................................................................................... 44 3.2 Hypothesis ....................................................................................................................................... 47 3.3 Materials and Methods ................................................................................................................... 47 3.3.1 Analytical Testing ......................................................................................................................... 47 3.3.2 Modeling Software ....................................................................................................................... 48 3.3.3 Pilot RO System ............................................................................................................................ 48 3.3.4 Scanning Electron Microscopy (SEM) .......................................................................................... 48 3.3.5 Energy Dispersive X-ray (EDX) ...................................................................................................... 49 3.3.6 Chromatic Elemental Imaging (CEI) ............................................................................................. 49 3.3.6 Fourier Transform Infared Technology (FTIR) .............................................................................. 49 3.3.7 Effervescing .................................................................................................................................. 50 3.4 Avista Autopsy and Analysis (membrane run with pH adjusted feed) ........................................... 50 3.4.1 Results of the Salt Passage, Fujiwara Analysis, Dye Testing ........................................................ 51 3.4.2 Results of the Microscopy Testing ............................................................................................... 52 3.4.3 Results of the Fourier Transform Infared Technology (FTIR) Analysis ......................................... 53 3.4.4 Results of Effervescing ................................................................................................................. 53 3.4.5 Cleaning Study ............................................................................................................................. 53 3.5 Confirmation of the Avista Microscopy Testing at the MSU SEM Facility ...................................... 55 3.5.1 SEM Images .................................................................................................................................. 56 3.5.2 EDX Mapping ................................................................................................................................ 58 3.6 SEM analysis (membrane run without pH adjusted feed) .............................................................. 61 3.6.1 SEM Images .................................................................................................................................. 62 3.6.2 EDX Mapping ................................................................................................................................ 63 Chapter 4 – Optimization ...................................................................................................................... 69 4.1 Permeate Production over Time ............................................................................................... 71 4.2 Flux over time ................................................................................................................................. 72 4.3 Pressure increase over time............................................................................................................ 74 4.4 Feed Quality .................................................................................................................................... 75 4.5 Permeate Quality ............................................................................................................................ 77 4.6 Concentrate Quality ........................................................................................................................ 78 4.7 Ammonia Balance ........................................................................................................................... 81 4.8 Membrane Cleaning ........................................................................................................................ 81 4.8.1 Cleaning Routines ........................................................................................................................ 83 4.8.2 Cleaning Frequency ...................................................................................................................... 84 4.8.3 Cleaning Costs .............................................................................................................................. 84 4.9 Economic evaluation ....................................................................................................................... 85 4.9.1 Electrical Costs ............................................................................................................................. 85 4.9.2 Capital Costs ................................................................................................................................. 85 4.9.3 Operating Costs ............................................................................................................................ 85 Chapter 5 – Conclusion and Future Recommendations ....................................................................... 86 REFERENCES .......................................................................................................................................... 88
vi
LIST OF TABLES
Table 1 - Calssification of hardness by carbonate concentration (MacAdam, 2004) ........................... 38 Table 2 - Air stripped water analytes. ................................................................................................... 46 Table 3 – Average analytical results of the 9 EDX scans (membrane run with pH adjustment) .......... 60 Table 4 - Average analytical results of the 9 EDX scans (membrane run without pH adjustment) ...... 67 Table 5 - Air stripped water analytes. ................................................................................................... 70 Table 6 - pH range and temperature limits for Filmetc membranes. (DOW Form No. 609-23010-0211) .............................................................................................................................................................. 82 Table 7 - List of chemical cleaning solutions used for the various foulant types. ................................ 82 Table 8 - Example of a cleaning cycle at Car-Min-Vu Dairy addressing organic and silicate fouling (high pH) and inorganic scale (low pH). ......................................................................................................... 83
vii
LIST OF FIGURES
Figure 1 - Poor waste management leads to manure or silage spills that can be extremely harmful to local aquatics. (Dohr, 2014) .................................................................................................................... 2 Figure 2 - Manure storage can be limited or expensive to construct. (Fulhage, 2012) .......................... 3 Figure 3 - Demonstration of manure applied to fields. (Patz, 2014) ...................................................... 3 Figure 4 - McLanahan Nutrient Separation System processing sand and solid separated dairy manure. ................................................................................................................................................................ 4 Figure 5 - Membrane types and their rejection limits. (McGowan, 2001) ............................................. 8 Figure 6 - Cut out of a reverse osmosis membrane ("Membrane Construction," 2015)...................... 15 Figure 7 - A demonstration of various fouling by adsorption, cake layer formation, pore blocking, and depth fouling (Adams, 2012). ............................................................................................................... 25 Figure 8 - Flux decline as a consequence of fouling for four experimental runs. ................................. 26 Figure 9 – Monosilicic acid (Ning, 2011) ............................................................................................... 30 Figure 10 - Solubility of silica from temperatures rangning from 0 to 80 ⁰C (Zuhl, 2013).................... 31 Figure 11 – Concentration of dissolved silica between pH of 2 to 11 (Amjad, 1997) ........................... 32 Figure 12 - This displays how dispersants act on compounds such as silica. (Demadis, 2004) ............ 33 Figure 13 - The structural relationships of the carbonate mineral (Keener, 2011) .............................. 35 Figure 14 - Schematic representation of crystallographic unit cells for (a) calcite and (b) isostructural dolomite, as well as (c) aragonite and (d) vaterite. (Xu, 2014) ............................................................. 36 Figure 15 - Solubility of the carbonate mineral (Moles) at a pH of 4 to 12 (Javid, 2011) ..................... 37 Figure 16 - Graph of non pH adjusted AS water (displayed over a time of 24 hours to see the drastic drop in permeate production) .............................................................................................................. 44 Figure 17 - SEM photo at x5000 revealed a granular foulant noticed upon the membrane surface ... 45 Figure 18 - Effervessing performed on a calcium carbonate scaled membrane (“Calcium Carbonate Scale,” 2015) ......................................................................................................................................... 50 Figure 19 - Display of the SEM image (left) and CEI (right) analysis showing foulants due to sulfur, silica, carbon, magnesium, and calcium ............................................................................................... 52 Figure 20 - FT-IR spectral image of foulant that was removed from membrane surface .................... 53 Figure 21 - Membrane prior to cleaning ............................................................................................... 54
viii
Figure 22 - Membrane after cleaning ................................................................................................... 55 Figure 23 - Foulant observed on the membrane at x500 magnification showing a mixture of organic material, silica, and minor traces of inorganics .................................................................................... 56 Figure 24 – Foulant observed on the membrane at x5000 magnification showing a mixture of organic material, silica, and minor traces of inorganics .................................................................................... 57 Figure 25 - Granular foulant observed on the membrane at x5000 magnification showing a mixture of organic material, silica, and minor traces of inorganics ................................................................... 57 Figure 26 – Image of the membrane sample (run at low pH conditions) used for EDX mapping ........ 58 Figure 27 - EDX mapping portraying large amounts of carbon, oxygen, silica, and sulfur. Also showing very minor amounts calcium ................................................................................................................ 59 Figure 28 - EDX mapping demonstrating very minor amounts of iron and copper .............................. 60 Figure 29 – The foulant inspected at x500 magnification revealing a mixture of organic and inorganic material ................................................................................................................................................. 62 Figure 30 –The x1500 magnification shows bolus’ representing inorganic calcium carbonate scale, and random white regions representing silicate scale upon a layer of organic foulant. ..................... 62 Figure 31 - The x5000 magnification shows the detailed inorganic calcium carbonate scale (shown as round bolus’). The image also shows the depth of the fouling. ........................................................... 63 Figure 32 - Image of the membrane sample (run without acid addition) used for EDX mapping ........ 64 Figure 33 - EDX mapping portraying large amounts of carbon, oxygen, and silica. Nitrogen, calcium, magnesium were also present in significant amounts. ........................................................................ 65 Figure 34 - EDX mapping displaying small amounts of sulfur, sodium, chlorine, postassium, and iron. .............................................................................................................................................................. 66 Figure 35 - Comparison of foulant upon the membrane run with acid addition versus the membrane run without acid addition. Displayed in the graph are the elements which are most likely to cause issues with fouling. ................................................................................................................................ 67 Figure 36 - Experimental design ........................................................................................................... 70 Figure 37 – Displays the comparison of flux decline as a negative slope at the various operating conditions using sulfuric, hydrochloric, or no acid at various pH ranges. ............................................ 72 Figure 38 - A display of the UF permeate versus air stripped water at pH 6.5. There is essentially no difference in performance. ................................................................................................................... 73 Figure 39 - The flux rate of the experiments performed at various pH ranges using sulfuric acid, hydrochloric acid, and no acid addition is displayed. At a pH of 6.5 there were replicate experiments performed for hydrochloric and sulfuric runs. ..................................................................................... 74
ix
Figure 40 - Pressure increase shown as a positive slope is displayed in the graph. The runs with
sulfuric, hydrochloric and, and no acid addition at various pH ranges is graphed. .............................. 75
Figure 41 - This graph displays the feed water quality in units of COD for the hydrochloric and sulfuric runs at the various pH ranges. .............................................................................................................. 76 Figure 42 - This graph displays the feed water quality in units of ammonia concentration (mg/L) for the hydrochloric and sulfuric runs at the various pH ranges. ............................................................... 76 Figure 43 - This graph displays the permeate quality in units of COD for the hydrochloric and sulfuric runs at the various pH ranges. .............................................................................................................. 77 Figure 44 - This graph displays the permeate quality in units of ammonia concentration (mg/L) for the hydrochloric and sulfuric runs at the various pH ranges. ............................................................... 78 Figure 45 - This graph displays the concentrate quality in units of COD for the hydrochloric and sulfuric runs at the various pH ranges. ................................................................................................. 80 Figure 46 - This graph displays the concentrate quality in units of ammonia concentration (mg/L) for the hydrochloric and sulfuric runs at the various pH ranges. ............................................................... 80
x
KEY TO ABBREVIATIONS
Abbreviation Full Word
RO Reverse Osmosis
SEM Scanning Electron Microscopy
EDX Energy Dispersive X-ray
CEI Chromatic Elemental Imaging
FTIR Fourier Transform Infrared Technology
GFD Gallons per square Foot per Day
GPD Gallons Per Day
UN United Nations
TDS Total Dissolved Solids
BWRO Brackish Water Reverse Osmosis
UF Ultrafiltration
NF Nanofiltration
SMP Soluble Microbial Products
NOM Natural Organic Matter
EfOM Effluent Organic Matter
1
Chapter 1 - Overview
1.1 Introduction
The use of RO has grown immensely over the past 40 years (Baker 2012). Reverse osmosis
provides a means of reclaiming purified water fairly inexpensively. One reason RO has become more
popular is due to the scarcity of fresh water in many areas such as dry lands, densely populated areas,
and isolated islands (Mohamed, 2004). According to the United Nations about 85% of the world lives in
regions with inadequate fresh water supply and 783 million people do not have access to clean water
(“Water Cooperation,” 2015). On Earth, water management has become a concern in growing
populations since less than 1% of water is available for human use (“Water Sense,” 2015). Forty states in
the USA expect water shortages within the next decade (“Water Sense,” 2015). The scarcity of water
within the USA is typically caused by increasing populations, increased industry, and climate change
(WWDR4, 2012). Clean water reclamation has become a reality and a priority in the USA and in the
world.
1.2 Problem Statement
Membrane technology is a fairly inexpensive way to reuse and reclaim water from waste water
applications. The cost to desalinate sea water and brackish water has dropped from $1 per 250 gallons
to around $0.5 per 250 gallons in 5 years (Henley, 2013). Water reclamation and nutrient recovery are
important aspects in today’s world to protect the fragile environment. Livestock agricultural is a
significant contributor to waste generation (USDA, 2006). Due to the need for more dairy and hog
operations, the production of wastes from hog and dairy processing, meat processing, and manure
management has increased (USDA, 2012). For example, large-scale dairy operations have increased from
564 farms in 1992 to 1807 in 2012, a 45% increase in 10 years (MacDonald, 2014). The shift to larger
2
dairy operations is driven by higher profits from larger heard sizes (MacDonald, 2014). Manure
management has become a significant issue as farmers struggle to store (see Figure 2), transport, and
land apply manure (see Figure 3) (USDA, 2006). Manure spills have had disastrous effects to humans and
wildlife (portrayed in Figure 1). An example of this is when an eight-acre hog-waste lagoon in North
Carolina burst, spilling 25 million gallons of manure into the New River, killing around 10 million fish and
shutting down shellfishing in 364,000 coastal acres in 1996 (NRDC, 2015). In some cases over excessive
land application of nutrient-rich manure has been performed as a last resort to get rid of the waste
(USDA, 2006). A great example from a recent study displayed that some larger dairy operations apply
manure to crops at rates that are three times greater than smaller farms (MacDonald, 2009). Often
times manure storage volume dictates the timing of manure land application. In turn, this can result in
excess manure application and increase the potential to negatively impact surface water due to runoff.
The World Health Organization (WHO) drinking water guideline is 10mg/L and the U.S. Geological Survey
found that 15% of shallow groundwater sampled below agricultural and urban areas had nitrate levels
higher than 10mg/L (Payal, 2000).
Figure 1 - Poor waste management leads to manure or silage spills that can be extremely harmful to
local aquatics. (Dohr, 2014)
3
Figure 2 - Manure storage can be limited or expensive to construct. (Fulhage, 2012)
Figure 3 - Demonstration of manure applied to fields. (Patz, 2014)
Nutrient separation systems have become of interest to prevent pollution and meet various
environmental standards. However, nutrient recovery has its own challenges as manure is a very
complex waste and should be studied under pilot projects, analyzed, and optimized. Only a few projects
have used membrane technology to process dairy manure waste streams. Some of these projects are
listed in the articles written by Masse and Gou, where they discuss challenges and lack of research in
4
identifying potential foulants, optimizing filtration systems for dairy manure waste, calculating the costs
associated with processing dairy manure through filtration systems. Masse and Gou expressed that
fouling of the membrane technology is a major issue in the dairy industry and should be researched to
determine the foulants involved. Fouling leads to a host of problems so the knowledge of the foulants
involved would decrease operational and maintenance costs, since there would be less irreversible
fouling, less down time for cleaning, less use of chemicals for cleaning, and less frequent membrane
replacement. One example of a comprehensive nutrient separation system was developed by the
McLanahan Corporation. Prior to nutrient separation, manure is sand and solid separated, leaving
behind a liquid manure slurry. This slurry is sent to an anaerobic digester coupled to an ultrafiltration
(UF) system. The phosphorus rich slurry is returned to the digester, where it is concentrated, and the
volatile ammonia in the ultrafiltered permeate is air stripped and stabilized by the absorber, as
ammonium sulfate. Finally, the RO system processes either air stripped water or non-air stripped water
to produce clean water and concentrate that is rich in potassium. A schematic of the process is shown in
Figure 4.
Figure 4 - McLanahan Nutrient Separation System processing sand and solid separated dairy manure.
5
1.3 Objective
Nutrient separation systems offer a treatment alternative to enhance the environmental
standing of large animal agriculture operations. As stated earlier, nutrient recovery provides a
mechanism to segregate nutrients and produce clean water but it has its own challenges. RO membrane
technology is one method used to produce water but is prone to fouling and scaling if improper
pretreatment and maintenance is not established. Specifically, RO membranes are particularly sensitive
to carbonate fouling, organic fouling, and silicate fouling in waste water applications dealing with
agriculture. In order to limit the fouling potential, tests for carbonate, organics, and silicates should be
performed. After analyzing the potential for fouling and scaling, pretreatment strategies can be put into
place, and optimization of the system can be performed so that nutrients are concentrated and clean
water is produced efficiently.
6
Chapter 2 – Literature Review
RO is currently used to purify water from sea water, brackish water, and waste water (Baker).
Approximately 70% of the Earth is covered in water and 97% of that coverage is sea water, so
desalination with use of RO is momentous (Perlman, 2014). In June 2011, 15,988 desalination plants
were producing 17.6 billion gallons of water per day, and supplying 300 million people worldwide with
that water (Henthorne, 2012). According to the article "Desalination industry enjoys growth spurt as
scarcity starts to bite," water retrieved from desalination increased to 20.7 billion gallons in 2013.
Reclamation of high quality water has also become prevalent in waste water applications, preventing
the pollution of groundwater and natural aquatics due to discharge of waste water. “RO membranes
have proven to successfully treat such waste water and provide water that exceeds reuse quality
requirements (Bartels, 2015).” Industrial and agricultural facilities have been able to treat waste water
using RO.
2.1 History
As the book “Reverse Osmosis Industrial Applicarions and Processes” by Jane Kucera states, the
earliest record of thin film semipermeable membranes was in 1748, discovered by Abbe Nollet, while
observing the phenomena of osmosis. Over the years this finding evolved, and by 1959, at the University
of Florida, C.E. Reid and E.J. Breton demonstrated that a cellulose acetate film could act to desalinate
water. Next was the commercialization of the cellulose acetate filters by optimizing the flux rate and
durability of the membrane. The flux rate is calculated from the amount of clean water produced per
day by the RO system. In 1960 Loeb and Srinivasa researched how to use this cellulose acetate
membrane under water pressure, and made RO commercially viable since it significantly improved flux.
The flux rated of the improved cellulose acetate membrane was about ten times greater than that of
other known membrane materials. Shortly then after, the first brackish water RO facility was
7
constructed, using wound filtration membranes, in Coalinga CA. The wound membranes became quite
popular and are currently used today, with some new improvements. Innovations keep growing and the
use of the spiral wound membranes exceeded expectations for desalinating sea water and brackish
water. This has been due to the modification of the membranes by the use of various polyamide or
cellulose membranes, different configurations of design within the membrane, and increased surface
area. Over time, enhancements allowed certain membranes to operate at higher pressure, from 1000
psi to 1200 psi. One of the current advancements is the nanotechnology membrane which has been
designed to reject 95% sodium chloride and 99.3% calcium chloride. Membrane technology continues to
evolve just as RO progresses and operates among other sources of water like industrial and municipal
waste streams. (Kucera, 2010)
2.2 Applications
Water reclamation by RO has become very popular especially due to the advancement in
membrane technology, less frequent membrane replacement, and prices of membranes have become
relatively inexpensive (Baker, 2012). The price to desalinate sea water and brackish water has been
reduced from $1 per 250 gallons to around $0.5 per 250 gallons in 5 years (Henley, 2013). RO is one of
the best water purification methods available today since it rejects most dissolved solids and suspended
solids by separating small solutes from water. The RO operation rejects material around 10-4 microns
(Figure 5). The rejected solids can collect on the membrane surface and foul the membrane.
Sometimes it is required to pretreat the feed water to the RO system with antiscalant and either
acid or base in order to minimize fouling and scaling of the membrane surface. Occasionally it is also
required to perform mandatory cleanings on the membranes, depending on the quality of the
concentrated water being treated. RO systems are primarily used in industrial settings but there are also
small-scale systems that have been used in homes, yachts, ocean liners, and remote regions. The
primary role of RO is to extract nutrients and other compounds from water, specifically sodium, out of
the feed stream to produce potable water (Kershner, 2008). As interest in RO grew, it has been applied
8
to other industries. The large diversity of applications have been able to reclaim and purify water
effectively and cost-efficiently. Some examples of where RO technology has proven to be effective is in
the treatment of municipal waste water and hazardous waste, extraction of specific compounds in the
food and beverage industry, and the retrieval of organic and inorganic materials from chemical
operations. Each field has its own set of challenges depending on the water being treated. The following
industries listed are currently using RO.
2.2.1 Desalination of Sea Water
RO technology was initially designed to separate sodium and other minerals from salt water. RO
technology is mostly used in this industry. Christopher Gasson from Global Water Intelligence stated
roughly 1% of the world population is dependent on desalination processes and by 2025 the UN
projected that 14% of the world’s population will be faced with scarcity of water ("Desalination industry
Figure 5 - Membrane types and their rejection limits. (McGowan, 2001)
9
enjoys growth spurt as scarcity starts to bite"). There are currently between 15,000 to 20,000
desalination plants worldwide producing water at a rate of more than 5.3 million gallons per day (GPD).
This shows the immense effectiveness of desalination and proves that continuous efforts have been
made to the advancements of this membrane technology.
2.2.2 Reverse Osmosis Treating Municipal Wastewater
RO systems processing municipal wastewater are typically located in regions that lack water
resources. Membrane treatment of municipal wastewater has proven to be a cost effective way of
reclaiming water. RO membranes have also proved that they are worthy in this field by significantly
reducing total dissolved solids, heavy metals, organic pollutants, viruses, bacteria, and other dissolved
contaminants. A few of the municipal RO plants include the 13.2 million GPD (gallons per day) plant in
West Basin, CA, the 10.6 million GPD plant in Singapore, and the 8.5 million GPD Bedok plant in
Singapore (Chilekar, Hydronautics). Some plants that process far more waste water than mentioned
before include the 71.3 million GPD plant in Orange County, California and the 100 million GPD plant for
Sulayabia, Kuwait (Bartels, Hydronautics). In the USA only 10% of water is used for drinking and cooking,
and the rest is flushed down the toilet or drain. California currently uses recycled water for toilet
flushing currently lowering its need for water by a quarter ("Indoor Water Use in the United States,”
2013). Also, San Diego is currently using recycled water from municipal waste water treatment plants
because it imports 85% of its drinking water from Northern California and the Colorado River, which are
currently in a drought crisis (Cho, 2011). These waste water RO plants demonstrate the importance of
recovering water and the acceptance that this technology has gained over the years.
10
2.2.3 Reverse Osmosis in Food Processing
Due to the disproportioned ratio of food production versus the growth of the world population,
nutrient recovery and water recovery has become very valuable. The application of membranes in the
food and beverage industry has increased dramatically since the 1980’s to recover and reuse as much
water as possible (Ganorkar, 2012). These industries include fruit and vegetable processing, animal
products, various beverages, sugar refining, and grain products (Ganorkar, 2012). Soybean processing
plants have been using nanofiltration-RO systems (NF) developed to recover water from soybean
soaking water (Guu, 1997). The article named “Reverse Osmosis System Cuts Food Plant’s Eco Footprint”
stated that a plant in Wisconsin recently started using RO to soften water for the boiling process and, in
turn, has also increased the efficiency of the boilers because of reduced alkalinity. The article also
mentioned that the treated boiler water is used in the canning process for carrots, green beans, and
potatoes. Low flow RO can also be used to treat bottle washing water producing drinkable water or
discharge water in the beverage industry (Mavrov, 2000). One of the beverage industries that utilizes
ultrafiltration and RO is the dairy industry. The dairy industry generates about 10 gallons of pollutant per
gallon of processed milk (Vourch, 2008). Additionally, it takes about 35 gallons of water to produce 1
cup of yogurt, 42 gallons to produce 1 scoop of ice cream, 50 gallons to produce two slices of cheese, 90
gallons to produce 1 cup of Greek yogurt, and 109 gallons of water to produce 1 stick of butter (Lurie,
2014). In these cases RO is highly regarded as a means to recycle water and concentrate nutrients.
2.2.4 Reverse Osmosis in Manure Management
In recent years manure management has become very important in the agricultural settings. A
few study’s in the year 2009 and 2011 found that the majority of large-scale dairies applied manure to
croplands at a rate of about 3 times more than small-scale farms (Macdonald, 2009). The loss of
phosphorus, nitrogen, and potassium to the environment during manure management is highly possible.
11
Normally manure is temporarily stored for a certain period of time (usually over the winter months and
when crops are growing) and then land applied as crop fertilizer. These techniques are typically suitable
for small-scale farms, or if the manure storage and application is agronomically correct and
environmentally friendly on large-scale farms (“Nutrient Recovery,” 2010). Sometimes this is not the
case. According to the website “Facts about Pollution from Livestock,” California identified that the
major source of nitrate pollution was from livestock agriculture, polluting over 100,000 square miles of
groundwater. The article also stated that in 1993, poor plant management practices at a dairy
contributed to the contamination of Milwaukee’s drinking water, killing over 100 people, and made
400,000 sick. Over application of manure leads to diminished crop yields and can lead to nitrogen
filtering into groundwater (Cogger, 2004). It is obvious that in recent years there has been greater
awareness in managing the dispersal of nutrients efficiently with minimal runoff (“Nutrient Recovery,”
2010). Producers are also interested in adding value to operations by generating electricity or placing a
value on concentrated nutrients (“Nutrient Recovery,” 2010). Mechanical wastewater treatment is one
way to concentrate the nutrients from manure. Ultrafiltration and RO are mechanical filtration methods
that can segregate nutrients such as phosphorus, potassium, and nitrogen. RO can also produce clean
water that may be suitable for direct discharge or land irrigation. The methods to recover nutrients and
produce clean water is a great way to prevent pollution due to livestock agriculture (“Nutrient
Recovery,” 2010).
With the innovative solutions to manure management come advanced challenges. Some
challenges are from the engineering design of systems, cost of systems, and the waste that needs to be
treated. In this study the manure is the waste that needs to be treated using membrane technology. The
greatest challenge is membrane fouling because the RO systems operate to produce clean water at the
molecular level. The fouling can be due to various elements that adhere to the membrane surface. RO is
still an attractive method for many industries since clean water is produced for reuse or is discharged.
12
Another benefit to using RO is the better environmental footprint RO systems leave behind by reducing
the discharge of waste into the environment, satisfying communities, meeting certain regulations, and
providing a positive influence to stay green and clean. There are also new innovations to RO technology
in order to address many of the challenges associated with the various applications.
2.3 Reverse Osmosis Knowledge Needed for Manure Management
As RO continues to expand, it has grown from primarily processing sea water, to reclaiming water in
the food and beverage industry, and has evolved to process waste water, or manure, in agriculture. RO
systems used in manure applications have been studied, but not entirely. Both Masse and Gou stated
that further controlled studies are required to develop a viable and economical technology for the use
of RO to process a downstream manure slurry (Masse, 2007, Gou, 2014). Masse and Gou also
mentioned that the following areas should be researched (Masse, 2007, Gou, 2014):
– Pretreatment:
A pretreatment strategy is needed in order to optimize system performance and keep costs low.
The effect of acid addition on transmembrane flux, reversible and irreversible fouling, cleaning
frequency, chemical requirements, permeate quality, and maximum volume reduction is also not
understood.
– Relationship between System Performance:
Various relationships with the following parameters are needed to better understand the
operation of the system to maximize performance. Masse and Gou recommended that the
following should be tabulated and compared: flux, fouling rate, concentrate characteristics,
volume reduction, permeate quality, pressure, and temperature.
13
– Major Types of Fouling
Masse and Gou emphasized the importance of understanding the fouling of the RO system.
Specifically, organic, inorganic, and biological fouling should be studied to understand what
compounds may be the primary cause for fouling. Once the components of fouling are
understood then a cleaning strategy and routine can be configured.
– Ammonia Volatilization
In order to document the volatilization of ammonia, a mass balance of ammonia across the RO
system is needed. Feed, concentrate, and permeate samples should be tested to observe the
amount of ammonia present in each stream and confirm the mass balance.
– Reuse of RO Permeate and RO Concentrate
In some cases the RO permeate may be suitable for disposal, otherwise it may be used
somewhere on the farm, for land irrigation, washing, cow cooling, used in a boiler system, or
possibly as animal drinking water. The concentrate would typically be stored and land applied
since it is rich in potassium and minor traces of phosphorus. It could also be dried in an
evaporator, to a solid form, extracting more water, for land application or commercial sale.
– Economic Evaluation
In order to understand the economics, all the capital costs and operational costs need to be
calculated and presented. This includes the acid addition, antiscalant addition, chemical cleaning
costs, costs associated with energy usage, replacement membranes, and the price of the entire
RO skid.
14
2.4 Reverse Osmosis Membranes
As previously stated, the RO system is a membrane based demineralization technique which
separates dissolved solids and suspended solids from solution for numerous source waters (Kucera,
2010). The suspended solids are typically known to foul the RO membrane (Roque, 2012). The most
common types are the thin film, cellulose acetate, composite membranes, which provide high rejection
and low operating pressure (Roque, 2012). If the source water has a high organic content, then the
cellulose acetate membranes are commonly recommended because they provide the least fouling rates
and shorter down time for cleaning (Roque, 2012).
RO membranes have evolved quite dramatically and new generations of membranes have been
produced offering adaptability to numerous applications (Roque, 2012). The hollow fiber modules used
to be quite popular for membrane filtration (Roque, 2012). These membranes are very sensitive to pH
change, pressure, and temperature (Roque, 2012). The hollow fiber membranes also became outdated
due to their high potential for fouling and scaling however, are sometimes used in applications with low
suspended solids content (Roque, 2012). More recently, the spiral wound membranes are preferred in
many fields of water treatment (Kucera 2010). The spiral wound RO membranes have multiple flat sheet
membrane leaves wrapped around a perforated permeate collection tube (Baker, 2004). Feed flows in
on one side of the membrane while the permeate passes the membrane on the other side and is
collected within the collection tube (see Figure 6). Each manufacturer has engineered various models for
different treatment applications.
15
Figure 6 - Cut out of a reverse osmosis membrane ("Membrane Construction," 2015).
RO membranes are typically made of cellulose acetate, cellulose diacetate, polysulfone,
polyethersulfone or polyamide material (Kucera, 2010). Cellulose acetate has been the most popular
(Kucera, 2010). All the membranes have a relatively smooth surface which potentially prevents fouling
(Kucera, 2010). Most of the membranes exhibit a neutral charge (Kucera, 2010). Depending on the
manufacturer, the maximum operating temperature is 113 degrees Celsius and the operating pH is
typically 4 to 7 (Kucera, 2010). Operating pressure is limited up to about 400 psig to prevent membrane
16
compaction (Kucera, 2010). The polyamide membranes were engineered to improve the performance of
the cellulose acetate membranes (Kucera, 2010). The chemical compositions of the membranes vary
between manufacturers and, depending on the process, one material may be favored over another to
Decomposed vegetable material cause well water to have a high concentration of dissolved
NOM. The major components of NOM are humic acids and fulvic acids.
2.7.8 Soluble Microbial Products (SMP) and Microbial Deposits
These complexes are found in biological wastewater treatment processes from substrate
metabolism. This fouling could also be due to bacterial slimes, fungi, molds, and other microbes.
Biocides are typically used to clean membranes and inhibit microbial growth with SMP and microbial
deposits.
In order to optimize an RO system the constituents and degrees of membrane fouling must be
understood. The RO system parameters for performance, cleaning protocols and pretreatment
strategies must be made to maximize production of high quality permeate.
2.8 Membrane Fouling in Manure Management
In the dairy industry, in order for a cow to produce one gallon of milk it produces two gallons of
manure (Dicktrell, 2014). Concentrating nutrients at milk processing plants with use of membrane
filtration has become very valuable and important and it is just as important to concentrate nutrients in
manure management to minimize pollution and agricultural runoff. The major nutrients are nitrogen,
phosphorus, potassium, calcium and magnesium, while the micronutrients are copper, iron, manganese,
and zinc (Yang, 2007). Organic material, in the form of humus substances, is also present. Additionally,
there are varying concentrations of boron, carbonate, and silica (Eriksson, 2001). The mineral content
mainly depends on the location of the farming operation, the feed ration, and the quality of water which
the cattle drink. The various compounds incorporated prior to downstream processing, specifically
related to membrane fouling should be considered. The processing of manure wastewater treatment is
23
quite similar to municipal wastewater treatment as similar treatment strategies are used, such as
anaerobic digestion, dissolved air flotation, clarification, centrifugation, ultrafiltration, microfiltration,
nanofiltration, and RO . Biological, organic, and inorganic fouling has been observed with membrane
technology processing manure wastewater (Masse, 2007). The biological constituents are usually
microbial matter, the organics are typically from humus substances or soluble microbial byproducts, and
the inorganic fouling is usually from carbonate or other inorganic constituents from the feed,
groundwater, or pretreatment. Just as brackish water, municipal wastewater, and milk processing
experience fouling or scaling from pretreatment, there may be additional compounds like aluminum,
iron, lime, sulfur, and polyelectrolytes from manure pretreatment that can cause scaling or fouling (Lin,
2013). There are numerous challenges from fouling that must be overcome when using RO. Once
optimized, the RO system would be able to operate efficiently at an acceptable flux rate and minimized
fouling rate. With nutrient recovery systems that incorporate polymers for coagulation before RO, there
may be major risks with irreversible fouling and deterioration of spiral wound membranes (Juang, 2001).
When using a nutrient recovery system that incorporates ultrafiltration and air stripping as
pretreatment to RO there is a low total solids content (about 0.7 wt %), no microbes or bacteria, lower
amount of volatile ammonia, and the only concern would be the compounds that could cause scaling or
fouling. The majority of these compounds are colloidal natural organic material, carbonate salts, and
possibly some elements from pretreatment.
2.9 Dynamics of Membrane Fouling
The fouling of the membrane system is influenced by the membrane type, quality of feed, and
fluid dynamics of the system (Adams, 2012). The most important concept of the fouling mechanism is
the concentration polarization.
24
2.9.1 Concentration Polarization
The dynamic accumulation of feed solids at the membrane surface due to the balanced
convective transport toward a membrane and the rate of diffusion away from the membrane is
concentration polarization (Cheryan, 1998). Balanced convective transport is the convection due to an
induced pressure gradient force (Frye 1913). The gel layer is a boundary which results from the
accumulation of the solids on the membrane surface. This occurs from supersaturation of the reject
(concentrate) that may result from increasing viscosity due to filtration forming the gel like boundary
layer (Adams, 2012). The gel layer then prevents the passage of permeate and osmotic pressure builds
up at the membrane surface acting against the trans membrane pressure (Adams, 2012). This is very
important in RO processes, but not as important in larger pore size systems such as ultrafiltration and
mediafiltration (Adams, 2012). Concentration polarization naturally occurs in membrane systems due to
the hydrodynamic conditions, and is not caused by the membrane itself (Marshall, 1993). Flux rate lost
to concentration polarization can be completely or partially restored when either the trans membrane
pressure is decreased, feed concentration is decreased, or cross-flow velocity in increased (Cheryan,
1998).
2.9.2 Fouling Mechanisms
Fouling occurs by four mechanisms after concentration polarization is in effect. These
mechanisms are adsorbtion, pore blocking, cake layer formation, and depth fouling (See Figure 7)
(Brans, 2004).
In dairy settings, proteins from milk production adsorb to polymeric, non-cellulosic membranes
under static conditions so some adsorption and fouling may occur well before concentration
polarization starts (Adams, 2012).
25
Figure 7 - A demonstration of various fouling by adsorption, cake layer formation, pore blocking, and
depth fouling (Adams, 2012).
On most occasions fouling is reversible by performing membrane cleaning, however, irreversible
fouling can sometimes occur and cleaning will not be able to restore full membrane performance.
Absorption occurs when foulant adheres to the membrane and may occur on the membrane surface or
within the pores (Adams, 2012). Absorbtion in the pores prevents passage of permeate and hence
reduces the flux (Adams, 2012). Cake layering is formed when foulants adsorb onto the surface of the
membrane or when particles agglomerate and bridge over sections of the porous membrane surface
(Adams, 2012). Compression of particulate that becomes caught in the bridging or piling of various
foulants can also occur adding additional layers that prevent and resist the passage of permeate
(Adams, 2012). Pore blocking occurs when particles that are larger than the pore become lodged at the
pore entrance (Adams, 2012). Another form of blocking is depth fouling and it occurs when a large
particle is forced deep into a pore through which it would not normally pass (Adams, 2012). This occurs
when there is an excessive trans-membrane pressure and therefore reduces membrane permeability
26
(Adams, 2012). Typically fouling can be cleaned from the membrane, however irreversibly bound
foulant, such as depth fouling, limits membrane performance and its lifespan (Renner, 1991).
2.9.3 Stages of Fouling
After concentration polarization and fouling, the resistance from these mechanisms increases
contributing to higher trans-membrane pressure (Fritsch, 2008). The concentration polarization and
fouling can be 10 to 50 times the resistance contributed by the membrane itself in the filtration process
(Hanemaaijer 1989). The trend of flux decline in cross-flow RO membrane processes is depicted in Figure
8:
Figure 8 - Flux decline as a consequence of fouling for four experimental runs.
Concentration polarization promotes fouling and is often called stage I flux decline (Marshall,
1993). Very early in RO process, during stage 1, the membrane flux rapidly drops, within seconds or
minutes. Immediate foulant adsorption also adds to this rapid flux decrease, as adsorption of protein to
the membrane surface occurs without concentration polarization (Tong, 1988). Since many polymeric
0
1
2
3
4
5
6
7
8
9
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Flu
x (G
FD)
Time (hrs)
27
varieties of membranes are deformable, membrane compaction may also be responsible for the
resistance of permeation and decrease in flux during the initial stage (Marshall, 1993). Stages II and III of
membrane systems flux decline are due to fouling. The drop in flux during stage II occurs by the initial
deposition of foulant onto the membrane and is a less dramatic decline compared to concentration
polarization (Marshall, 1995). Finally, stage III is an asymptotic decline and can be due to the additional
deposition and compaction of the foulant layer (Marshall, 1995). After solutes and colloidal particles
become adsorbed onto the membrane surface, cake layers or monolayers are formed (Belfort, 2004).
These monolayers start overlapping and form multilayers, then the multilayers are compacted under the
system’s trans-membrane pressure (Belfort, 2004). The RO process is quite simple, however, the fouling
mechanisms and concentration polarization can be complex and it is recommended that the stages of
fouling are understood when trying to evaluate the effectiveness of a membrane filtration system.
2.10 Foulants of interest
The specific foulants of interest are silica, carbonate, and natural organic material because they
are assumed to cause complications with this specific project, and are known to be challenging for many
waste water applications that use RO. The applications section, Section 2.2, provided great insight about
the various compounds that could be of interest. Silica was chosen because it is found in manure at
concentrations around 100 mg/L and is a common concern in many RO applications (Amjad, 2008).
Natural organic matter is also of interest since manure organics are commonly found in manure at high
concentrations. Carbonate scale is also problematic in many RO settings, but can be addressed by
pretreatment much easier than silica and natural organic material. Combined, these inorganic and
organic foulants can drastically reduce the efficiency of RO systems.
28
2.10.1 Silica
Silica is found in sand and plants, can constitute 2-5% of dry leaf mass in plant matter (Massey,
Ennos and Hartley, 2006) and is indigestible to many animals (Saha, 2010). Silicic acid is one of the major
constituents in the soil solution which becomes deposited in plant roots (Epstein, 1994). Typically, the
organic matter in flooded soil causes a higher mobility of silica due to the ferric hydrous oxides that also
releases the silicic acid (Kabata-Pendias , 2001). The silica is taken up by plants as an essential
micronutrient to support the shoot of the plant (Epstein, 1994). “Plants deprived of Si are often weaker
structurally and more prone to abnormalities of growth, development and reproduction and it is the
only nutrient which is not detrimental when collected in excess” (Epstein, 1999). Silica is also very
prevalent in sand. Sands contain small quantities of heavy rock forming minerals (“Sand,” 2015). Quartz,
a form of silica, is the most common element in all types of sand because of silica’s abundance in rocks
(“sand,” 2015). Quartz can be defined as the complete dehydrated form of silicic acid (Ning, 2003). As
rocks erode they produce silica enriched sand (“sand,” 2015). The silicic acid in the sand then
depolymerizes by re-hydration when contacted with a water stream and becomes a soluble form of
silicic acid (Ning, 2003). The concentrations may vary depending on the exposure to geologic material.
The forms of the silica mineral (silicon dioxide) are quartz, tridymite, cristobalite, coesite,
stishovite, lechatelierite, and chalcedony (“silica mineral,” 2015). Tridymite, cristobalite, and the
hydrous silica mineral opal are quite uncommon (“silica mineral,” 2015). Vitreous silica, coesite, and
stishovite, are even more uncommon than the latter, and are considered extremely rare (“silica
mineral,” 2015). Silica is the group of minerals composed of silicon and oxygen, which are the most
copious elements in the earth's crust (“silica mineral,” 2015). About 28% of the Earth’s crust contains
silica minerals (“silica mineral,” 2015). Silica is commonly found in well water, as the vitreous form, and
can reach values of 60 parts per million (Peairs, 2007). Water collected near volcanic activity or oil fields
can have silica concentrations as high as 300 parts per million (Meyers, 1975). Enormous amounts of
29
silica are carried away in solution from weathering rocks and soils every year (“silica mineral,” 2015).
Silica is also prevalent in surface waters that contain biological activity (Peairs, 2007). Certain algal
microorganisms, like diatoms, integrate reactive silica to construct a protective shell made of silicon
dioxide crystals (Peairs, 2007). When the algal microorganisms decompose, the silica is released into the
environment as reactive silica (Peairs, 2007). The interaction of biological organisms also creates the
colloidal silicates which are prevalent in most surface waters (Peairs, 2007). There are high
concentrations of reactive silica in well water and some surface water since the water is in contact with
dissolving rock (Peairs, 2007). Silica is prevalent on the earth and is present in essentially all forms of
water.
2.10.1.1 Silica’s Chemical Properties
“Silica has historically created problems for water treatment because of its stability as an un-
ionized compound, making it difficult to remove using ion-exchange processes, and is in fact one of the
least preferred anions to treat (Peairs, 2007).” The bonding of silicon is very similar to that of carbon,
with its four valence electrons, however, it is an inert element (Myshli︠a︡ eva, 1974). Crystalline silica has a
very similar composition to silicon dioxide where 46.75 % by weight is silicon and 53.25 % by weight is
oxygen (“silica mineral,” 2015). Crystalline silica has a very low solubility (6 mg/L) in water compared to
amorphous forms of silica (100 to 140 mg/L) (Amjad, 2008). All silicate compounds, except for stishovite,
are crystallographic structures (“silica mineral,” 2015). These structures are three-dimensional arrays of
linked tetrahedrons, each consisting of a silicon atom coordinated by four oxygen atoms (“silica
mineral,” 2015). The tetrahedrons contain silicon-oxygen bond distances of 1.16±0.002 Å and the main
differences between the silicate compounds is the geometry of the tetrahedral linkages (“silica mineral,”
2015). The stishovite structures are octahedrons where silicon atoms bond with 6 oxygen atoms (“silica
mineral,” 2015). Silicon has a high affinity to oxygen and it is quite problematic to separate silicon
30
dioxide, SiO2 (“silica mineral,” 2015). All of the silicate compounds have a specific gravity around 2 to
2.7 g/L (“silica mineral,” 2015). The solubility increases with increased temperature (“silica mineral,”
2015). The solubility also increases in the presence of �� groups and � �� (“silica mineral,” 2015).
Quartz is the least soluble out of all the silica compounds. In pure water, at 25°C, the solubility of quartz
is 6 parts per million (“silica mineral,” 2015).
The compound SiO2 bonds with itself and can form a tetrahedral, crystalline forming a lattice, or
can be in a non-crystalline form and is classified as reactive (dissolved), colloidal, or suspended
particulate (Ning, 2011). Reactive silica is much smaller than the colloidal form and can be distinguished
as monosilicic acid, disilicic acid and polysilicic acid (Ning, 2011). The reactive form dissolves in water
and forms monosilicic acid (Figure 10) (Ning, 2011).
Figure 9 – Monosilicic acid (Ning, 2011)
Monosilicic acid is unionized at natural pH levels, 10% ionized at a pH of 8.5, and 50% ionized at
a pH range of 9 to 10 (Peairs, 2007). The silicate colloids are generally thought to be either silicon that
has polymerized with numerous elements of silicon dioxide, or silicon that formed bonds with organic
compounds or other complex inorganic compounds (Peairs, 2007). When silica forms anhydrides from
the reaction of silicic acid and organic compounds, silica scale is increased and therefore limits recovery
rates of filtration systems (Ning, 2011). At higher temperatures the compounds which bond to silica are
usually aluminum and calcium oxide structures, increasing the solubility of silica (Peairs, 2007). Silica can
also interact with compounds such as nitrogen, sulfur, phosphorus, aluminum, iron, and some halides
31
(Kabata-Pendias, 2001). When silica is present in acidic conditions within the soil, silica and phosphate
ions can form insoluble precipitates (Kabata-Pendias, 2001). Finally, particulate silica is larger in size and
mostly comprised of sand, like quartz, or suspended solids in water (Ning, 2011).
Silica and metal silicate-based salts are the most problematic foulants in industrial wastewater
systems. When silica forms anhydrides from the reaction of silicic acid and organic compounds, silica
scale is increased and, therefore, limits recovery rates of filtration systems (Ning, 2011). Notice the
trough in the solubility graph, Figure 12, between the pH range of 7 to 7.5 where the solubility is
extremely low. Also, as temperature increases the solubility of silica increases as seen in Figure 11.
Figure 10 - Solubility of silica from temperatures rangning from 0 to 80 ⁰C (Zuhl, 2013)
32
Figure 11 – Concentration of dissolved silica between pH of 2 to 11 (Amjad, 1997)
2.10.1.2 Silica Scaling and Fouling
Fouling due to silica will occur at low pH ranges and low temperatures (see Figures 11 and 12).
“Silica solubility is well known to be both pH and temperature dependent (Amjad, 1997).” At low pH,
and if metal ions are present (ions such as aluminum, iron, calcium, and magnesium), then the fouling
due to silica is exacerbated (Amjad, 2008). Silica becomes more soluble at high pH, so acidifying does not
help. However, it is highly soluble at higher temperatures, above 80 degrees Fahrenheit. The silica
solubility is also limited if certain compounds such as sodium chloride or magnesium chloride are
present (Hamrouni, 2001). This can be challenging when trying to solubilize silica. Silicate fouling can
occur due to condensation of monomeric silicic acid on solid substrates containing hydroxyl groups (-
OH), polymerization of silicic acid or colloidal deposition, and biogenic amorphous silica by living
33
organisms (Amjad, 2009). Fouling due to precipitation occurs when monomeric silica, also called silicic
acid, polymerizes on the RO membrane surface (Amjad, 2008). Particulate fouling occurs when colloids
accumulate during the polymerization process onto the RO equipment and membranes. Depending on
the structure of the silica layer, transport of solutes can be convective or diffusive. If the layer is colloidal
or particulate then solutes transport through the layer dominated by convective flow, but if the silica
layer is made up of polymerized silica then the solutes transport via diffusion (Amjad, 1997).
Prevention of silica fouling is best addressed by using dispersants and high temperature (Peairs,
2007). “The results of a pilot RO study showed that deposition of silica and magnesium silicate on
membrane surface can be prevented by the use of a polymeric dispersant (Amjad, 1997).” Antiscalants
containing dispersants can prevent silica polymerization and scatters fine particles of amorphous silica
once they have formed (Gill, 1990). Dispersion is defined as a method to finely divide a substance in
solution (“Terminology of Polymers,” 2011). The appropriate dosing of antiscalant is necessary and can
be calculated by using antiscalant simulation software, such as the Avista Advisor from Avista
technologies, which contains models constructed by Avista Technologies. Figure 13 displays how
dispersion works, it prevents the agglomeration and adhesion of silicon dioxide onto the membrane
surface.
Figure 12 - This displays how dispersants act on compounds such as silica. (Demadis, 2004)
34
2.10.1.3 Carbonate
Carbonate is derived from either carbonic acid or carbon dioxide. Carbonates can also be
classified as inorganic or organic forms ("carbonate,” 2015). Inorganic carbonates are made from
carbonic acid salts (H2CO3) which contain the carbonate ion (CO ��) and metals such as calcium and
sodium ("carbonate,” 2015). The hard shells of many marine invertebrates are made from inorganic
carbonate ("carbonate,” 2015). Another inorganic form is the carbonate mineral and is the most widely
distributed within the Earth’s crust ("carbonate mineral,” 2015). Organic carbonates, esters, contain the
carbon group ethyl (C2H5), which takes the place of hydrogen atoms of carbonic acid ("carbonate,”
2015). Fifty percent of the carbonate and bicarbonate salts that exist in natural water can be due to
weathering (Chapman 1996). The concentration of carbonate and bicarbonate in surface waters is
generally less than 500 mg/L and more commonly less than 25 mg/L (Chapman 1996). Groundwater is
sometimes more alkaline with concentrations of carbonate or bicarbonate of up to 10 mg/L while
surface waters generally contain lesser amounts of carbonate because their pH rarely exceeds 9
(Chapman 1996). Carbonate salts are also contained in some of the livestock’s feed, but most salts
entrained in manure are from the water supply (Johnson, 2006).
There are over eighty known forms of the carbonate mineral, which are constituents of certain
rocks, and the most common varieties of carbonate are calcite, dolomite, and aragonite (“carbonate
mineral,” 2015). Calcite is the principal mineral of limestone and marble (“carbonate mineral,” 2015).
However, when there is an excess amount of dolomite in limestone, the rock is typically named
dolomite (“carbonate mineral,” 2015). Aragonite is found in calcareous skeletons, within the shells of
organisms and also found in some sediment (“carbonate mineral,” 2015). Some other carbonate
minerals are found in metal ore and the most common are siderite, rhodochrosite, strontianite,
smithsonite, witherite and cerussite (“carbonate mineral,” 2015).
35
2.10.1.4 Carbonate’s Chemical Properties
The carbonate ion has a trigonal symmetry which allows most carbonate minerals to form
crystal like structures (“carbonate mineral,” 2015). The carbon atom within the carbonate mineral is
centrally located and surrounded by oxygen atoms (“carbonate mineral,” 2015). The anion within the
carbonate mineral then bonds with compounds such as aluminum, barium, calcium, copper, iron, lead,
manganese, sodium, uranium, and zinc (“carbonate mineral,” 2015). The structure of carbonate changes
as protons are removed and changes from carbonic acid to bicarbonate to the carbonate ion (Figure 14).
The structural relationship is also dependent on pH, shown in Figure 17. This is important to understand
since when pH adjusting the carbonate ion can be in the form of primarily one structure or 2 structures,
and the structure can be fully or partially soluble at various pH ranges (Figure 16).
Figure 13 - The structural relationships of the carbonate mineral (Keener, 2011)
Typically most of the rocks that contain carbonate are either structured by calcite or aragonite
(“carbonate mineral,” 2015). The calcite is structured very similarly to that of sodium chloride in a
rhomohedral form. In this case the sodium and chloride groups become calcium and carbonate groups,
respectively. Within the calcite structure CO3 groups lay in parallel and horizontal layers and CO3 groups
in adjacent layers point in opposite directions. The calcium atoms bond with 6 oxygen atoms and the
36
calcium atom is distributed one each from three CO3 groups in a layer above and three from CO3 groups
in a layer below. Dolomites structure is similar to that of calcite, except that there is an extra
magnesium, and a lower symmetry. The aragonite structure is orthorhombic and like the calcite
structure, the cation in the aragonite structure is surrounded by 6 carbonate groups, however, they are
rotated about an axis perpendicular to their plane and the cation is matched with nine oxygen atoms
rather than six. Bicarbonate is formed when half the acidic hydrogen in carbonate is replaced by a metal,
such as calcium (MacAdam, 2004). Hydrated carbonates, bicarbonates, and compound carbonates
containing other anions in addition to carbonate are some other forms of carbonate minerals
(MacAdam, 2004). Figure 15 illustrates the various structures carbonate can form when clustered
together. These structures are calcite and isostructural dolomite, as well as aragonite and vaterite
(Figure 15)
Figure 14 - Schematic representation of crystallographic unit cells for (a) calcite and (b) isostructural
dolomite, as well as (c) aragonite and (d) vaterite. (Xu, 2014)
The natural alkalinity of groundwater, earths buffering system, is comprised primarily of bicarbonate,
37
carbonate, and hydroxide ions. The buffering occurs when small doses of strong acid, for example acid rain, react with the alkalinity in water. The acid converts carbonate to bicarbonate, converts bicarbonate to carbon dioxide, and this all occurs with a minor change in the pH of the water. During buffering the dissolved CO2 may react with water to form a weak acid called carbonic acid. In the pH range of 4.2 to 4.5, or 8.2 to 8.4, carbon dioxide and bicarbonate are balanced. The alkalinity is in the form of carbon dioxide at a pH of 4.2 to 4.5 while at a pH of 8.2 to 8.4 most alkalinity is in the bicarbonate form with not much carbon dioxide present. At a pH around 8.2 or 9.6 there can be a balance of carbon dioxide and bicarbonate. Typically, at a pH of 9.6, no carbon dioxide or bicarbonate is present, and the majority of alkalinity is carbonate. When the pH of water is above 9.6 alkalinity occurs due to the presence of hydroxyls, the presence of the hydroxide ion. Natural water sources can have a pH in the range of 6 to 8.4 and the presence of hydroxides is predominantly due to human impacts. Alkalinity can be measured using chemical indicators and reported as M-Alkalinity and P-Alkalinity in terms of “ppm as calcium.” M- Alkalinity is normally called the Total Alkalinity and measures the amount of carbonate, bicarbonate, and hydroxide present. The P-Alkalinity measures the concentration of hydroxyl and carbonate alkalinity. (Bates, 2015)
Figure 15 - Solubility of the carbonate mineral (Moles) at a pH of 4 to 12 (Javid, 2011)
38
2.10.1.5 Carbonate Scale
Before delving into the details of carbonate scale, harness of water should be understood.
Hardness occurs when divalent metal cations bond to anions, like carbonate and sulfate, to form a
precipitate (MacAdam, 2004). Total hardness is usually quantified by the concentration of magnesium
and calcium cations (MacAdam, 2004). Temporary harness is measured by the concentration of
bicarbonate and carbonate salts. The classification of hardness is shown in Table 1.
Table 1 - Calssification of hardness by carbonate concentration (MacAdam, 2004)
Concentration (���� as ���� ) Degree of hardness
0-50 Soft
50-100 Moderately soft
Table 1 (cont’d)
100-150 Slightly hard
150-250 Moderately hard
250-350 Hard
350+ Excessively hard
In order to limit carbonate scale the RO unit must be run under acidic conditions, below a pH of seven
(Mullin, 2001). Calcium carbonate crystallization occurs in three phases called supersaturation,
nucleation, and crystal growth. During supersaturation, there is an induction period where the first
nucleus is formed (Mullin, 2001). Post supersaturation, calcium ions and carbonate ions begin to cluster
and form a stable nuclei during the nucleation step (Mullin, 2001). This cluster continues to cultivate and
becomes the crystal growth phase (Mullin, 2001). In some cases, temperature influences crystallization
39
where low temperature increases heterogeneous precipitation and high temperature induces
homogeneous precipitation (Mullin, 2001). Scaling occurs when the foulant is transferred from bulk
solution and binds to surface. The scale can increase strength by recrystallizing and also withstands
erosion. At high pH ranges bicarbonate becomes carbonate, which allows for more carbonate fouling to
exist (Andritsos, 1999).
2.10.2 Organics (EfOM)
Effluent organic matter (EfOM) is a form of wastewater effluent and when introduced to the RO
membrance can contribute to fouling (Barker, 2000). The organic compounds associated with EfOM are
polysaccharides, proteins, aminosugars, nucleic acids, humic and fluvic acids, organic acids, and other
cell compounds (Barker, 2000). These intricate compounds are classified into two groups: soluble
Masse and Gou stressed that pretreatment is important since membrane scaling and fouling is
the major concern for RO systems. A high fouling rate leads to higher than normal pressure drop, higher
operating pressure, and reduced permeate generation. Pretreatment is crucial in order to maintain
operation of a RO system in a challenging waste water environment.
If there is an inadequate pretreatment routine, an RO system will require frequent cleaning
intervals and membrane life will be short. Determining a cost effective pretreatment will reduce the
intensity of cleaning and may enhance membrane life. The pretreatment that was chosen for this
71
project chemical based where acid and antiscalant were batch fed to a tank prior to processing through
the RO system. The acids that were compared were concentrated hydrochloric acid and concentrated
sulfuric acid. As discussed earlier, the acids are used to chemically reduce, remove, destroy or inhibit
bacteria, hardness scale, and oxidizing agents. The antiscalant used in the process is called Vitec 4000,
which is primarily used for waters containing high concentrations of silica but also prevents carbonate
and sulfate scale. As stated earlier, antiscalants work to keep supersaturated salts in solution, change
the shape of crystals to move through the membrane, and impart a high negative charge to various
crystals to prevent propagation. For this particular project the acid addition and use of antiscalants was
the most cost effective pretreatment technique.
Sulfuric acid is typically cheaper than hydrochloric but both were compared to test their
effectiveness on the reduction of membrane fouling. In theory, when using sulfuric acid, the formation
of sulfate scale increases, such as barium sulfate, calcium sulfate, and strontium sulfate. Also, according
to the Avista Advisor and ROSA modeling software less hydrochloric is needed to lower pH compared to
sulfuric acid. Since less HCl is necessary for pH adjustment it may seem beneficial to use HCl instead of
H2SO4 depending on the economics and if HCl can perform better than sulfuric acid on reducing
membrane fouling.
4.1 Permeate Production over Time
The following graph (Figure 39) displays the permeate production and flux over time at the various
feed pH ranges, pH = 5.5, 6.5, 7.5, and no pH adjustment (pH = 8). The feed pH of 6.5 was the best
performer, since the slope of permeate production (minor axis in Figure 39) was the least and produced
the most permeate in a 24 hour period, compared to the rest. The performance of the system was
determined by the amount of permeate generated in a 24 hour period. Since the pH of 6.5 was the best
overall, replicate experiments were executed to confirm its performance. Also, surprisingly, the RO
system performance was pretty much the same using hydrochloric acid verses sulfuric acid. Sulfuric acid
72
was the preferred acid to use since it is much cheaper to use than hydrochloric acid and it allows the RO
system to work efficiently (more permeate generated in 24 hours). The use of acid also enhanced the
longevity of the run, indicating there was less membrane fouling when acid was used and resulting in
less frequent membrane cleaning. With sulfuric acid used for pretreatment, the UF permeate at a pH of
6.5 processed through the RO system produced a similar amount of RO permeate as the air stripped
water processed through the RO system at a pH of 6.5, but had a higher ammonia concentration,
therefore only air stripper runs were performed for this study.
Figure 37 – Displays the comparison of flux decline as a negative slope at the various operating
conditions using sulfuric, hydrochloric, or no acid at various pH ranges.
4.2 Flux over time
The flux is the gallons of permeate produced per square foot of membrane per day. On this
specific RO system, the total area of all the membranes combined was 340 square feet. The declining
flux rate per hour had the same trend as the permeate generation rate per hour. According to Figure 41,
the pH at 7.5 showed to have a high flux but a poor permeate quality according to ammonia levels
0
0.05
0.1
0.15
0.2
0.25
0.3
pH = 5.5 pH = 6.5 pH = 7.5 pH = 8
Slo
pe
(N
eg
ati
ve
)
Sulfuric
Hydrochloric
Non pH Adjusted
73
(Figure 45). At the pH of 5.5 the flux rate was also high, however, the permeate quality was not very
good, according to the COD values (Figure 45). Again, the experiment using air stripped water pH
adjusted with sulfuric acid at a pH of 6.5 proved to be the best contender, since it had the second
highest flux rate, and better permeate quality. The same repeated trials with air stripped water at pH 6.5
were executed to confirm the performance using sulfuric and hydrochloric acid. There were five
experiments using air stripped water adjusted with sulfuric to a pH of 6.5, two experiments using air
stripped water adjusted with hydrochloric to a pH of 6.5, and two experiments using UF permeate
adjusted with sulfuric to a pH of 6.5. The error bars, standard deviation, for air stripped water at pH 6.5
using sulfuric was 1.67 and 0.43 when using hydrochloric (Figure 40). Also, the UF permeate processed
through the RO system had the same or similar flux as the air stripped water processed through the RO
system at the pH of 6.5 using sulfuric acid (Figure 40). The standard deviation for UF permeate was 0.57
and 1.17 for air stripped water at a pH of 6.5 using sulfuric acid (Figure 40).
Figure 38 - A display of the UF permeate versus air stripped water at pH 6.5. There is essentially no
difference in performance.
0
1
2
3
4
5
6
UF Permeate Air Stripped Water
Flu
x R
ate
(G
FD
)
74
Figure 39 - The flux rate of the experiments performed at various pH ranges using sulfuric acid,
hydrochloric acid, and no acid addition is displayed. At a pH of 6.5 there were replicate experiments
performed for hydrochloric and sulfuric runs.
4.3 Pressure increase over time
The pressure increase over time was monitored to investigate the fouling rate. As pressure
increases the membrane becomes scaled over and pores become blocked. When the pores become
blocked, the pressure increases and the permeate generation rate decreases. The minor axis in Figure 42
is the slope of the pressure as it increases over time. The slope is positive and the greater the slope the
greater the overall pressure resulting from a high fouling rate. The run with sulfuric acid and a pH of 6.5
proved to be one of the better runs since the slope was less than one, while the rest of the trials had a
slope above one (Figure 42). The run with a feed pH of 7.5 by use of hydrochloric acid was at a slope of
one and very similar to the run using sulfuric acid at a pH of 6.5, however the permeate quality at the pH
7.5 was lower than that of the sulfuric run at pH 6.5 (Figure 46). For the most part the other trials had a
pressure increase above one. The most significant pressure increase was when no pH adjustment was
0
1
2
3
4
5
6
pH = 5.5 pH = 6.5 pH = 7.5 pH = 8
Flu
x R
ate
(G
FD
)
Sulfuric
Hydrochloric
Non pH Adjusted
75
made, at a slope of 14 in Figure 42. The rest of the trials were comparable, but the run with sulfuric at a
pH of 6.5 was best.
Figure 40 - Pressure increase shown as a positive slope is displayed in the graph. The runs with
sulfuric, hydrochloric and, and no acid addition at various pH ranges is graphed.
4.4 Feed Quality
The feed stream to the RO system has high COD and high ammonia values. This can be seen in
the graphs below (Figure 43 and 44). Figure 43 also shows very little differences for COD and ammonia
values between the 3 pH adjustment targets (pH of 5.5, 6.5, and 7.5). The reason the ammonia
concentration for the permeate was higher at pH 5.5 in this graph is because at the time, the air stripper
system was being optimized and unfortunately this run had produced a higher ammonia value. The
ammonia value for numerous runs using sulfuric acid at a pH of 6.5 had shown to be one of the best
since it was consistently lower, around 335mg/L (Figure 44). Replicates were gathered at the pH of 6.5.
Three samples at a pH of 6.5 using sulfuric acid and two samples at a pH of 6.5 using hydrochloric acid
were used for COD and Ammonia in Figure 43 and 44. For COD, the standard deviation error bars were
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
pH = 5.5 pH = 6.5 pH = 7.5 pH = 8
Slo
pe Sulfuric
Hydrochloric
Non pH Adjusted
76
378 for use with sulfuric and 71 for use with hydrochloric. The ammonia concentration had SD error bars
that were 78 for sulfuric and 127 for hydrochloric.
Figure 41 - This graph displays the feed water quality in units of COD for the hydrochloric and sulfuric
runs at the various pH ranges.
Figure 42 - This graph displays the feed water quality in units of ammonia concentration (mg/L) for the
hydrochloric and sulfuric runs at the various pH ranges.
0
500
1000
1500
2000
2500
3000
3500
4000
4500
pH = 5.5 pH = 6.5 pH = 7.5
CO
D (
mg
/L)
Sulfuric
Hydrochloric
0
100
200
300
400
500
600
700
800
pH = 5.5 pH = 6.5 pH = 7.5
Am
mo
nia
Co
nce
ntr
ati
on
(m
g/L
)
Sulfuric
Hydrochloric
77
4.5 Permeate Quality
Again, there was a total of three trials using sulfuric acid and two trials using hydrochloric at a
pH of 6.5, since the trials at 6.5 performed the best (produced the most permeate) at the lowest cost
and best permeate quality. The permeate quality was determined by COD and ammonia values. For this
reason the COD and ammonia values were gathered, averaged and graphed (See Figure 45 and 46
below). The RO permeate had a fairly low COD, ~20mg/L, and low ammonia concentration, ~60mg/L,
when using sulfuric at a pH of 6.5. When using hydrochloric, similar results are displayed however it
costs more to pH adjust so it is not recommended. In Figure 45 the standard deviation at the pH of 6.5
using sulfuric was 29 and 16 when using hydrochloric. The standard deviation in Figure 46 is 52 for
sulfuric and 32 for hydrochloric. The feed at a pH of 7.5 using sulfuric and hydrochloric had produced the
same COD result (between 20 and 30mg/L) as the pH of 6.5 however the ammonia value was quite high
(between 160 and 180mg/L). At a pH of 5.5 the permeate COD was over 100mg/L. If the permeate
meets EPA and other government standards it could be reused for cow drinking, barn washing, direct
discharge or possibly other uses at the facility.
Figure 43 - This graph displays the permeate quality in units of COD for the hydrochloric and sulfuric
runs at the various pH ranges.
0
10
20
30
40
50
60
70
80
90
100
110
pH = 5.5 pH = 6.5 pH = 7.5
CO
D (
mg
/L)
Sulfuric
Hydrochloric
78
Figure 44 - This graph displays the permeate quality in units of ammonia concentration (mg/L) for the
hydrochloric and sulfuric runs at the various pH ranges.
4.6 Concentrate Quality
The concentrate quality is listed below and contains the remainder amount of ammonia, COD,
and retains essentially all of the remaining nutrients, especially potassium. The COD and ammonia data
values are portrayed in Figure 47 and 48. The results show that for feed with sulfuric dosing, the COD is
very high, 12000 mg/L, in concentrate at the lowest pH, 9833 mg/L at the pH 6.5, and 7400 mg/L for the
pH 7.5. The results show that the feed adjusted with H2SO4 had concentrate ammonia values highest at
the pH of 5.5 (2100 mg/L) and values about the same at the pH 6.5 (1757 mg/L) and 7.5 (1200 mg/L).
The standard deviation for the COD data at pH 6.5 was 1401 for sulfuric and 566 for hydrochloric. The
feed adjusted with hydrochloric revealed that the concentrate COD was about the same at pH levels of
5.5, 6.5, and 7.5 (7800 to 8800 mg/L). The concentrate of the feed adjusted with hydrochloric showed
that the ammonia values were lowest at pH of 5.5 (1027 mg/L) and highest at the pH of 7.5, 1400mg/L.
The pH level of 6.5 using hydrochloric had an ammonia value of 1115 mg/L. The standard deviation of
0
20
40
60
80
100
120
140
160
180
200
pH = 5.5 pH = 6.5 pH = 7.5
Am
mo
nia
Co
nce
ntr
ati
on
(m
g/L
)
Sulfuric
Hydrochloric
79
the ammonia concentrations at pH 6.5 was 1205 for sulfuric and 262 for hydrochloric. Finally, the
concentrate is high in potassium which is an essential nutrient for various crops. This is a highly
concentrated form of potassium and could be sold or used on the farm as a fertilizer.
80
Figure 45 - This graph displays the concentrate quality in units of COD for the hydrochloric and sulfuric
runs at the various pH ranges.
Figure 46 - This graph displays the concentrate quality in units of ammonia concentration (mg/L) for
the hydrochloric and sulfuric runs at the various pH ranges.
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
11000
12000
13000
pH = 5.5 pH = 6.5 pH = 7.5
CO
D (
mg
/L)
Sulfuric
Hydrochloric
0
500
1000
1500
2000
2500
3000
3500
pH = 5.5 pH = 6.5 pH = 7.5
Am
mo
nia
Co
nce
ntr
ati
on
(m
g/L
)
Sulfuric
Hydrochloric
81
4.7 Ammonia Balance
The average ammonia balance was calculated from two of the runs performed using sulfuric acid for pH
adjustment at a pH of 6.5. The mass balance, of ammonia, revealed that both of the runs had a
consistent 90% ammonia removal efficiency.
4.8 Membrane Cleaning
Cleaning should be performed when there is evidence of fouling. Membrane cleaning is
important to maintain permeability of the RO process. Fouling is evident when normalized permeate
flow has decreased less than 10% from the starting flow, normalized permeate quality has decreased
less than 10% from the starting quantity, and normalized pressure drop between feed and concentrate
increases about 15% from the starting pressure (Hydronautics TSB107.21, 2011). It is also important to
clean the RO membranes when they are lightly fouled rather than heavily fouled. In some cases,
frequent cleaning is mandatory and membrane replacement will most likely occur every year or every
few years. In certain applications membranes can be cleaned once a day and in more extreme cases
sometimes a 30 second cleaning every 30 minutes (DOW Form No. 609-00306-800, 2000). Chemical
cleaning of the RO membranes removes deposits by chemical reactions including hydrolysis, peptization,
solubilization, dispersion, chelation, sequestering, and suspending (Tragardh, 1989). This restores the
capacity as well as separation characteristics of the system. Selecting the correct cleaning products and
cleaning procedure greatly depends on the assortment of foulants present upon the membrane (DOW
Form No. 609-00306-800, 2000). Chemical cleaning with acid and base reacts with deposits to dissolve
foulants and keep the foulants dispersed and in solution. Low pH is used to clean inorganic material and
high pH is used to clean organic material that adheres to the membrane. DOW recommends cleaning
their Filmetec membranes with an alkaline cleaning solution first. Most Filmtec polyamide membranes
82
can be safely cleaned at the pH range of 1 to 12 according to appropriate temperature range shown in
Table 6. (DOW Form No. 609-23010-0211)
Table 6 - pH range and temperature limits for Filmetc membranes. (DOW Form No. 609-23010-0211)
The process of adding alkaline chemicals elevates the solubility and negative charge on the organic
based foulant (Ang, 2009). DOW Filmtec’s “Cleaning and Sanitation” technical bulletin mentioned that
using EDTA (disodium ethylenediaminetetraacetate) and SDS (sodium dodecyl sulfate) can increase the
cleaning efficiency on the RO membranes. EDTA is a metal chleating agent that is used to remove
organic material that contain divalent cations (Ang, 2009). Anionic surfactants, such as SDS, have
hydrophilic and hydrophobic groups, are semi soluble in organic and aqueous solvents, and are used to
remove foulants by solubilizing macromolecules by forming micelles around them (Ang, 2009). Inorganic
scale due to metal oxides, carbonates and sulfates are normally cleaned by using acids such as
hydrochloric acid or citric acid. Acids will dissolve precipitants that are lodged on the membrane surface
(Arnal, 2011). Below is a chart, Table 7, which displays chemical agents that are responsible of removing
various foulants.
Table 7 - List of chemical cleaning solutions used for the various foulant types.
Fouling due to Chemical Solutions
Carbonate scale HCl, EDTA, SDS
Sulfate scale HCl, EDTA
Phosphate scale HCl
Metal oxide scale Citric acid, HCl, Na2SO4
Silicate scale NaOH
83
Table 7 (cont’d)
Colloids NaOH, EDTA, SDS
NOM (Natural Organic Material) NaOH, EDTA, SDS
Biofilms NaOH, EDTA, SDS, Disinfectants
4.8.1 Cleaning Routines
A general cleaning sequence includes numerous stages such as product removal by rinsing with
clean water, chemical cleaning in a series of steps, and finally rinsing again with water. Filmtec
recommends using chlorine-free water such as RO permeate or deionized water for rinsing. Softened
water is recommended if RO permeate or deionized water is not available. Depending on the foulants,
the chemical cleaning may involve multiple rinses with high or low pH and circulation with various
cleaning solutions. Rinsing and cleaning should be pumped at a low flow rate, around 40 psi, and certain
solutions may be heated as per membrane specs. Soaking is optional and effective but may conflict with
mandatory operational time. To determine a successful cleaning, the permeate generation rate should
be higher and the concentrate pressure lower than those used during normal operation (Tragardh,
1989). The permeate rate should be the same or close to the manufacturer specifications under certain
operating conditions to determine whether or not a cleaning cycle has been effective. An example of a
typical cleaning at Car-Min-Vu dairy, using the four membrane pilot system, addressing mostly organic
and silica fouling is shown in Table 8.
Table 8 - Example of a cleaning cycle at Car-Min-Vu Dairy addressing organic and silicate fouling (high
pH) and inorganic scale (low pH).
Method (high pH)
Chemical
Vol.
(gal)
Water
(gal)
Prev. Soln.
(%)
pH Temp.
(°F)
Circ.
time
Clean water rinse - - 50 - 7 ~70 -
High pH rinse 50% NaOH 0.09 50 - 12 95 -
High pH rinse and circulate 50% NaOH 0.08 50 20 12 95 45
High pH rinse and circulate 50% NaOH 0.08 50 20 12 95 45
Avista P111 rinse and circulate 50% NaOH Avista P111
0.03 3lbs
50 - 12 95 45
Clean water rinse - - 50 - 7 ~70 -
84
Table 8 (cont’d)
Clean water flux test - - - - 7 85 -
Method (low pH)
Chemical
Vol.
(mL)
Water
(gal)
Prev. Soln.
(%)
pH Temp.
(°F)
Circ.
time
Clean water rinse - - 50 - 7 ~70 -
Low pH rinse 50% HCl 50 - 1.5 70 -
High pH rinse and circulate 50% HCl 0.08 50 20 12 95 45
High pH rinse and circulate 50% HCl 0.08 50 20 12 95 45
Avista P111 rinse and circulate 50% HCl Avista P303
0.03 3lbs
50 - 12 95 45
Clean water rinse - - 50 - 7 ~70 -
Clean water flux test - - - - 7 85 -
4.8.2 Cleaning Frequency
Without pH adjustment the pilot RO system would have to be cleaned about every four hours
due to a rapid flux decline (Table 8). With use of acid and antiscalant the RO system does not need to be
cleaned as often. From the trials performed on this pilot system, and using the preferred pH of 6.5 using
sulfuric acid, the recommended cleaning frequency of the membrane system is once a day. This could
potentially be extended to one and a half days or two days if the trials were to be replicated for 48 hour
runs.
4.8.3 Cleaning Costs
The chemical costs to clean the RO is the largest cost operation. For this pilot membrane system
50% NaOH, 32% HCl, and Avista ROclean P111 would be needed (in bulk). In order to clean a 20 element
system every day it would cost $2.82 for 50% NaOH, $7.44 for 32%HCl, and $42.68 for the Avista P111.
The total daily cost to clean the RO is $94.17.
85
4.9 Economic evaluation
4.9.1 Electrical Costs
The average electrical cost is $0.10/kWh, according to the US Department of Energy. According
to the ROSA simulation software the amount of energy required to process 1000 gallons of feed is
2.28kWh. The total electrical cost to run a 20 element system is approximately $0.2/kgal or $0.0002/gal.
4.9.2 Capital Costs
Membranes cost about $550 per module according to the local distributor (Purchase
Advantage). The total cost, including the membrane cost, for an RO skid is $3100 per module. Each
module has a filtration area of 400 square feet. A 20 element system would process dirty water over
8000 square feet of membrane and the capital cost would be roughly $62000.
4.9.3 Operating Costs
Essentially, there are three main aspects to the operating costs. These are primarily chemical
costs. One chemical cost is for pretreatment, the other chemical cost is for daily cleaning. Finally, a labor
cost of $25 an hour at 0.35 full time equivalents (FTE) is necessary, for cleaning and maintenance. If
hydrochloric was used to pretreat the feed stream from a pH around 8 to a pH of 6.5 the cost would be
$0.0119/gallon versus using sulfuric which would cost $0.0032/gallon to adjust air stripped water. The
total cost for cleaning chemicals is $94.17 a day.
86
Chapter 5 – Conclusion and Future Recommendations
The RO systems have come a long way from initially specializing in rejecting salt and minerals
from salt water to treating some of the most challenging waste streams such as municipal waste and
agricultural waste (manure slurry). This project determined what types of elements from a manure
slurry may pose problems due to fouling and scaling on the spiral wound membranes of the RO system.
The main foulants are organic material (such as humic acid and fulvic acid) and inorganic constituents
(carbonates and silica). The organic fouling can be inhibited by using antiscalants that include chelating
agents such as EDTA salts and the inorganic fouling due to carbonates and silica can be inhibited by pH
modification and addition of antiscalant. The FTIR and SEM – EDX imaging proved that organic fouling is
always present. The imaging analysis showed that when no pH adjustment is made there is more
inorganic foulant present. Less inorganic foulant is present when there is pH adjustment, using acid. By
using these tools the RO system was optimized to fit this system, and sulfuric acid was the
recommended acid used for pretreatment. By adjusting the pH of the feed to 6.5 and completing the
pretreatment stage with proper use of antiscalant, and performing at its best, the RO system can
effectively produce permeate at around 1GPM and a flux rate of about 4.3GFD (Figure 41) . Ammonia
and COD was tracked throughout the study and also confirmed that the most stable and best quality
permeate was delivered using sulfuric acid to pH adjust air stripped water at a pH of 6.5. Finally the
cleaning study determined the appropriate timing for cleaning as well as chemicals required to clean the
RO system. Concentrated NaOH, HCl, and Avista ROClean P111 were used to clean the system with a
recommended cleaning protocol. The costs to use each chemical for cleaning and pH adjustment were
gathered and included in the operational costs. The capital costs were also calculated. Also, the
permeate and concentrate samples were frequently tested during the course of the study to determine
the ammonia mass balance and the permeate quality from this particular RO system. At a pH of 6.5 the
ammonia removal efficiency was 90%.
87
A highly replicated study is recommended to determine the exact elements, especially organics,
that are the culprits for the RO fouling. By knowing which organic elements are fouling the RO system
the pretreatment strategy could be modified to enhance the efficiency, permeate production, of the
system. This is particularly difficult since every location is different and will have diverse analytical,
water chemistry data, depending on the water and food provided to the cows for drinking and eating.
There may also be an influence with the bedding used for the cows, whether it be sand (rich in silica),
manure solids, or fresh hay. Further analytical testing should be performed once more systems are in
place to determine the differences in water quality prior to the RO system and determine what analytes
may cause fouling issues to the RO system. A highly replicated study to compare various cost effective
pretreatment strategies is also recommended.
The RO system can run without supervision, with use of a proper automated pretreatment
routine, and needs mandatory cleaning every day. Further research could also be performed to verify
that a 48 hour run or maybe a 32 hour run is sustainable to save money on operational, cleaning, and
maintenance costs. The RO system is an extremely important aspect in the nutrient separation system
to ensure the highest quality water is permeated and reused somewhere in the dairy. RO is a step
forward in closing the sustainability loop and bringing water back to the beginning of the dairy industry
cycle.
88
REFERENCES
89
REFERENCES
"The 4 C’s of Manure Spill Response." Michigan State University Extension. Web. 2 July 2014. "Calcium Carbonate Scale." Avista. Web. 02 August 2015. "Carbonate." Encyclopedia Britannica Inc. Web. 29 June 2015. "Carbonate Mineral." Encyclopædia Britannica Inc. Web. 29 June 2015. "Chromatic Elemental Imaging." Color Imaging Fundamentals and Applications (2008): n. pag. Avista.
Avista, Aug. 2013. Web. "Desalination Industry Enjoys Growth Spurt as Scarcity Starts to Bite." Global Water Intel. Web. 01 June
2015. "Facts About Pollution from Livestock Farms." NRDC 21 February, 2013. Web. 17 August 2015. "Indoor Water Use in the United States." EPA. United States Environmental Protection Agency, 2013.
Web. "Membrane Construction." AvistaTechnologies. Web. 22 July 2015. "Nutrient Recovery: State of Knowledge." WERF. December 2010. Web. 17 August 2015. "Reverse Osmosis System Cuts Food Plant's Eco Footprint." Environmental Leader RSS 27 August, 2013.
Web. 21 July 2015. "Scale Control." DOW, Filmtec August, 2000. Web. "Silica Mineral." Encyclopedia Britannica Inc. Web. 16 June 2015. "Surface Analysis." Encyclopædia Britannica Inc. Web. 01 August 2015. "Ultrafiltration Water Purification, Microfiltration Membranes." Meco. Web. 22 July 2015. "Water Sense - Statistics & Facts." Environmental Protection Agency. Web. 01 June 2015. Adams, Michael Corey. "Examination of Methods to Reduce Membrane Fouling During Dairy
Microfiltration and Ultrafiltration." Cornell University, 2012. Print. Alexander, GB. "The Preparation of Monosilicic Acid." Journal of the American Chemical Society 75.12
(1953): 2887-88. Print. Alghoul, MA, et al. "Review of Brackish Water Reverse Osmosis (Bwro) System Designs." Renewable and
Sustainable Energy Reviews 13.9 (2009): 2661-67. Print.
90
Amjad, Z, and RW Zuhl. "An Evaluation of Silica Scale Control Additives for Industrial Water Systems." NACE International, New Orleans (2008). Print.
Amjad, Z. Silica Control in Industrial Water Systems with a New Polymeric Dispersant. AWT 2009 Annual Convention, Hollywood, FL. 2009. Print.
A New Antifoulant for Controlling Silica Fouling in Reverse Osmosis Systems. IDA World Congress on Desalination and Water Reuse, Madrid. 1997. Print.
Andritsos, N, and AJ Karabelas. "The Influence of Particulates on Caco3 Scale Formation." Journal of heat
transfer 121.1 (1999): 225-27. Print. Ang, Wui Seng, Sangyoup Lee, and Menachem Elimelech. "Chemical and Physical Aspects of Cleaning of
Atteberry, Johnathan. "How Scanning Electron Microscopes Work" 21 April 2009.
HowStuffWorks.com Baker, Richard W. Membrane Technology and Applications. West Sussex, United Kingdom: John Wiley
and Sons Ltd, 2012. Print. Bartels, Craig R. "Reverse Osmosis Membranes for Wastewater Reclamation." Nitto Hydronautics. Web.
2015 2 June. Bates, Wayne. "Ro Water Chemistry." Hydronautics 2015. Web. Batista-Garcia, Valerie. "Treating Brackish Groundwater in Texas: A Comparison of Reverse Osmosis and
Nanofiltration." US Department of the Interior May 2015. WebTreating Brackish Groundwater in Texas: A Comparison of Reverse Osmosis and Nanofiltration
Belfort, Georges, Robert H Davis, and Andrew L Zydney. "The Behavior of Suspensions and
Macromolecular Solutions in Crossflow Microfiltration." Journal of Membrane Science 96.1 (1994): 1-58. Print.
Boyles, Wayne. "The science of chemical oxygen demand." Technical information series, Booklet 9 (1997) Brans, GBPW, et al. "Membrane Fractionation of Milk: State of the Art and Challenges." Journal of
Membrane Science 243.1 (2004): 263-72. Print. Chapman, Deborah V, and World Health Organization. "Water Quality Assessments: A Guide to the Use
of Biota, Sediments and Water in Environmental Monitoring." (1996). Print. Cheryan, Munir. Ultrafiltration and Microfiltration Handbook. CRC press, 1998. Print. Chilekar, Satish. "Production of High Purity Water for Semi-Conductor Industry Using City Sewage by
Public Utility Board of Singapore: A Case Study." (2015). Print.
91
Cho, Renee. "From Wastewater to Drinking Water." State of the Planet. Earth Institure, Columbia University, 4 Apr. 2011. Web.
Cogger, Craig. "Manure on Your Farm: Asset or Liability?" Livestock and Poultry Environmental
Stewardship (n.d.): n. pag. Washington State University. 2 Dec. 2004. Web. Demadis, Konstantinos D, and Eleftheria Neofotistou. "Inhibition and Growth Control of Colloidal Silica:
Designed Chemical Approaches." Materials performance 43.4 (2004): 38-42. Print. Dickrell, Jim. "From Liquid Manure to Drinkable Water." AgWeb, Dairy Today 09 June, 2014. Web. 11
September 2015. Epstein, Emanuel. "The Anomaly of Silicon in Plant Biology." Proceedings of the National Academy of
Sciences 91.1 (1994): 11-17. Print. "Silicon." Annual review of plant biology 50.1 (1999): 641-64. Print. Eriksson, Jan. Concentrations of 61 Trace Elements in Sewage Sludge, Farmyard Manure, Mineral
Fertiliser, Precipitation and in Oil and Crops. Vol. 5159: Swedish Environmental Protection Agency Stockholm, Sweden, 2001. Print.
Fritsch, JACIM, and CI Moraru. "Development and Optimization of a Carbon Dioxide-Aided Cold
Microfiltration Process for the Physical Removal of Microorganisms and Somatic Cells from Skim Milk." Journal of dairy science 91.10 (2008): 3744-60. Print.
Fulhage, Charles D, Donald L Pfost, and John Feistner. Storage Tanks for Liquid Livestock Manure.
University of Missouri--Columbia, Extension Division., 2002. Print. Ganorkar, Pravin, Anil Nandane, and Ajay Tapre. "Reverse Osmosis for Fruit Juice Concentration–a
Review." Research & Reviews: Journal of Food Science & Technology 1.1 (2012). Print. Greenlee, Lauren F, et al. "Reverse Osmosis Desalination: Water Sources, Technology, and Today's
Challenges." Water research 43.9 (2009): 2317-48. Print. Griffiths, P. ; de Hasseth, J.A. "Fourier Transform Infrared Spectrometry " Wiley-Blackwell 18 May 2007.
Web. 01 August 2015. Guo, Xuejun, and Xin Jin. "Purification of Uf-Treated Anaerobically Digested Manure Wastewater by
Two-Pass Reverse Osmosis." Desalination and Water Treatment 52.16-18 (2014): 3027-34. Print. Guu, Yuan-Kuang, Chiu-Hsia Chiu, and Jing-Kun Young. "Processing of Soybean Soaking Water with a Nf-
Ro Membrane System and Lactic Acid Fermentation of Retained Solutes." Journal of Agricultural
and Food Chemistry 45.10 (1997): 4096-100. Print. Hamrouni, B, and M Dhahbi. "Analytical Aspects of Silica in Saline Water—Application to Desalination of
Brackish Waters." Desalination 136.1 (2001): 225-32. Print. Henley, Will. "The New Water Technologies That Could Save the Planet." The Guardian (2013). Print.
92
Hong, Seungkwan, and Menachem Elimelech. "Chemical and Physical Aspects of Natural Organic Matter (Nom) Fouling of Nanofiltration Membranes." Journal of membrane science 132.2 (1997): 159-81. Print.
Hydronautics. "Chemical Pretreatment for Ro and Nf." 111. Nitto October 2013. Web.
A Comparison of Conventional Treatment Methodsand Vsep, a Vibrating Membrane Filtration System. Proceedings of the El Paso Desalination Conference, El Paso Texas. 2006. Print.
Juang, Ruey-Shin, and Chwei-Huann Chiou. "Feasibility of the Use of Polymer-Assisted Membrane
Filtration for Brackish Water Softening." Journal of Membrane Science 187.1 (2001): 119-27. Print.
Kabata, Alina, and H Pendias. "Trace Elements in Soils and Plants." CRC, Washington, DC (2001). Print. King, Hobart. "The "Acid Test" for Carbonate Minerals and Carbonate Rocks." Geology.com. N.p., n.d.
Web. 19 Apr. 2015. Kershner, Kate. "How Reverse Osmosis Works." HowStuffWorks.com (2008). Print. Khan, Muhammad Tariq, et al. "How Different Is the Composition of the Fouling Layer of Wastewater
Reuse and Seawater Desalination Ro Membranes?" Water research 59 (2014): 271-82. Print. Koo, T, YJ Lee, and R Sheikholeslami. "Silica Fouling and Cleaning of Reverse Osmosis Membranes."
Desalination 139.1 (2001): 43-56. Print. Kucera, J. "Reverse Osmosis Design, Processes, and Applications for Engineers, 2010." Co-published by
John Wiley & Sons, Inc. Hoboken, New Jersey, and Scrivener Publishing LLC, Salem,
Massachusetts. Print. Kucera, Jane. "Properly Apply Reverse Osmosis." Chemical engineering progress 93.2 (1997). Print. Lenntech. Foulants and Cleaning Procedures for Composite Polyamide RO Membrane Elements (ESPA,
ESNA, CPA, LFC, NANO and SWC) Http://www.lenntech.com/Data-sheets/tsb107-L (2011): n. pag. Lenntech. Nitto Hydronautics, Oct. 2011. Web.
Lenntech. "Scaling and Antiscalants." Web. 30 August 2015. Lin, Hongjun, et al. "A Review on Anaerobic Membrane Bioreactors: Applications, Membrane Fouling
and Future Perspectives." Desalination 314 (2013): 169-88. Print. Lurie, Julia. "It Takes How Much Water to Make Greek Yogurt?!" Mother Jones 10 March 2014. Web. 21
July 2015. MacDonald, James, and Doris Newton. "Milk Production Continues Shifting to Large-Scale Farms."
Amber Waves (2014). 1E.
93
MacDonald, J.M., and W. D. McBride. "The Transformation of U.S. Livestock Agriculture: Scale, Efficency, and Risks." U.S. Department of Agriculture, Economic Research Service (2009). Electronic Information Bulletin No.EIB43.
MacAdam, Jitka, and Simon A Parsons. "Calcium Carbonate Scale Formation and Control." Re/Views in
Environmental Science & Bio/Technology 3.2 (2004): 159-69. Print. Marshall, AD, PA Munro, and G Trägårdh. "The Effect of Protein Fouling in Microfiltration and
Ultrafiltration on Permeate Flux, Protein Retention and Selectivity: A Literature Review." Desalination 91.1 (1993): 65-108. Print.
Masse, L, DI Masse, and Y Pellerin. "The Use of Membranes for the Treatment of Manure: A Critical
Literature Review." Biosystems engineering 98.4 (2007): 371-80. Print. Massey, Fergus P., A. Roland Ennos, and Sue E. Hartley. "Silica in Grasses as a Defence against Insect
Herbivores: Contrasting Effects on Folivores and a Phloem Feeder." Journal of Animal Ecology
75.2 (2006): 595-603. Print. Mavrov, V, and E Bélières. "Reduction of Water Consumption and Wastewater Quantities in the Food
Industry by Water Recycling Using Membrane Processes." Desalination 131.1 (2000): 75-86. Print.
McGowan, Wes. "Water Processing, 3rd Ed." Water Quality Association 2000. Web. Meyers, Peter. "Behavior of Silica in Ion Exchange and Other Systems." IWC 99 (1975): 64. Print. Mullin, John William. Crystallization. Butterworth-Heinemann, 2001. Print. Myshli︠a︡ eva, Lidii︠a︡ Vasil′evna, and Valencn Vasil′evich Krasnoshchekov. Analytical Chemistry of Silicon.
Halsted Press, 1974. Print. Ning, Robert Y. "Discussion of Silica Speciation, Fouling, Control and Maximum Reduction." Desalination
151.1 (2003): 67-73. Print. Ning, Robert Y, and Thomas L Troyer. "Colloidal Fouling of Ro Membranes Following Mf/Uf in the
Reclamation of Municipal Wastewater." Desalination 208.1 (2007): 232-37. Print. Okazaki, Minoru, et al. "Water Recycling Using Sequential Membrane Treatment in the Electronics
Industry." Desalination 131.1 (2000): 65-73. Print. Peairs, David. "Silica over-Saturation, Precipitation, Prevention and Remediation in Hot Water Systems."
Cal Water (2007). Print. Pinnau, I, and BD Freeman. "Formation and Modification of Polymeric Membranes: Overview."
Membrane Formation and Modification 744 (2000): 1-22. Print.
94
Rabiller-Baudry, M, et al. "Coupling of Sem-Edx and Ftir-Atr to (Quantitatively) Investigate Organic Fouling on Porous Organic Composite Membranes." Chapter in Current Microscopy
Contributions to Advances in Science and Technology, Formatex, Badajoz, Spain (2012). Print. Railsback, L Bruce. "Some Fundamentals of Mineralogy and Geochemistry." On-line book, quoted from:
www. gly. uga. edu/railsback (2006). Print. Renner, Edmund, and MH Abd-El-Salam. Application of Ultrafiltration in the Dairy Industry. Elsevier
Science Publishers Ltd., 1991. Print. Roque, Jennifer C. "Evaluation of an on-Line Device to Monitor Scale Formation in a Brackish Water
Reverse Osmosis Membrane Process." University of Central Florida Orlando, Florida, 2012. Print. Sangyoup Lee, Wui Seng Ang, Menachem Elimelech, Fouling of reverse osmosis membranes by
hydrophilic organic matter: implications for water reuse, Desalination, Volume 187, Issues 1–3, 5 February 2006, Pages 313-321, ISSN 0011-9164
Sampat, Payal. "Groundwater Shock." World Watch 13.1 (2000): 10-22. Print. Slomkowski, Stanislaw, et al. "Terminology of Polymers and Polymerization Processes in Dispersed
Systems (Iupac Recommendations 2011)." Pure and Applied Chemistry 83.12 (2011): 2229-59. Print.
Thomas, Robert. "A beginner’s guide to ICP-MS." Spectroscopy 16.4 (2001): 38-42. Tong, PS, DM Barbano, and MA Rudan. "Characterization of Proteinaceous Membrane Foulants and Flux
Decline During the Early Stages of Whole Milk Ultrafiltration." Journal of Dairy Science 71.3 (1988): 604-12. Print.
UNICEF. "Managing Water under Uncertainty and Risk, the United Nations World Water Development
Report 4, Un Water Reports, World Water Assessment Programme." UNESCO, Paris, France, 2012. Print.
USDA Natural Resources Conservation Service 12 December 2006. Web. "Facts About Pollution from Livestock Farms." USDA National Agricultural Statistics Service. 2012 Census
of Agriculture. Web. 30 August 2015. Visser, J, and Th JM Jeurnink. "Fouling of Heat Exchangers in the Dairy Industry." Experimental Thermal
and Fluid Science 14.4 (1997): 407-24. Print. Vourch, Mickael, et al. "Treatment of Dairy Industry Wastewater by Reverse Osmosis for Water Reuse."
Desalination 219.1 (2008): 190-202. Print. Waksman, Selman A. "Humus Origin, Chemical Composition, and Importance in Nature." Soil Science
41.5 (1936): 395. Print.
95
Wilf, Mark, and Steven Alt. "Application of Low Fouling Ro Membrane Elements for Reclamation of Municipal Wastewater." Desalination 132.1 (2000): 11-19. Print.
Williams, M. "A Review of Wastewater Treatment by Reverse Osmosis." EET Corporation and Williams
Engineering Services Company Inc (2003). Print. Xu, Ben, and Kristin M Poduska. "Linking Crystal Structure with Temperature-Sensitive Vibrational
Modes in Calcium Carbonate Minerals." Physical Chemistry Chemical Physics 16.33 (2014): 17634-39. Print.
Yang, J, et al. "Mechanistic Evidence and Efficiency of the Cr (Vi) Reduction in Water by Different Sources
of Zerovalent Irons." Water Science & Technology 55.1-2 (2007): 197-202. Print. Zuhl, Robert W, and Zahid Amjad. "10 Solution Chemistry Impact on Silica Polymerization by Inhibitors."
Mineral Scales in Biological and Industrial Systems (2013): 173. Print.