RECOVERY, TREATMENT, AND RECYCLING OF INDUSTRIAL WASTEWATER by KRISTIE LEA WITTER, B.S. A THESIS IN CIVIL ENGINEERING Submitted to the Graduate Faculty of Texas Tech University in Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE IN CIVIL ENGINEERING Approved Accep;ted December, 1997
97
Embed
RECOVERY, TREATMENT, AND RECYCLING A THESIS the ...
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
RECOVERY, TREATMENT, AND RECYCLING
OF INDUSTRIAL WASTEWATER
by
KRISTIE LEA WITTER, B.S.
A THESIS
IN
CIVIL ENGINEERING
Submitted to the Graduate Faculty of Texas Tech University in
Partial Fulfillment of the Requirements for
the Degree of
MASTER OF SCIENCE
IN
CIVIL ENGINEERING
Approved
Accep;ted
December, 1997
Ac g06 ^^''''' 13 I o^q 7 ACKNOWLEDGMENTS
C^lp' ^ I am indebted greatly to the members of my committee for their support and
direction of this thesis and for their most valuable criticism and academic nourishment.
Dr. Raghu Narayan, Dr. John Borrelli, and Dr. Ralph H. Ramsey III. A special thanks
to Dr. Richard Tock for his aid in the final examination of this research. For their
academic expertise, advice and direction, I wish to express my appreciation to
Dr. Tony Mollhagen and Dr. Alex Gilman.
A special thanks is extended to all the employees at Texas Instruments in
Lubbock, Texas, for without whose help this project could not have been completed.
I particularly wish to acknowledge Fernando Alvarez-Lara, Galen Kunka, Yimin
Chiou, Cindy RufiF, and Gerald Hector for their support above and beyond what was
requested. In addition, I wish include Ms. Shannon Reed in my acknowledgments for
the inner strength and fiiendship that she gave me during some of the most difficult
and best of times. It has made all the difference in my life!
Finally, I wish to thank each of my fiiends and family members for their
patience, understanding, and support throughout the entire course of this project. My
feelings are best stated by I Corinthians 12:26, "and if one member suffers, all
members suffer with it; or if one member is honored, all the members rejoice with it."
This has been particularly true throughout the course of my academic studies. For
their support during the trying times, I wish to honor and rejoice with them now.
Finally, I wish to give thanks to God, for it is through Him that all things are possible!
ii
TABLE OF CONTENTS
ACKNOWLEDGMENTS ii
LIST OF TABLES v
LIST OF FIGURES vii
CHAPTER I. ESTTRODUCTION 1
1.1 Purpose 2 1.2 Objectives 5 1.3 Case Study (Overview) 5
II. REVTEW OF LITERATURE 7 2.1 Treatment Options 7 2.2 Precipitation 8
Where: Qp = volumetric flow of water, M8= mass flow rate of contaminants (salts), Kw and Kg = membrane permeability coefficients (for water and salts respectively), Wp and Sp = water and sah permeabilhy coefiQcients, AP = hydrauUc pressure difference across the membrane. An = osmotic pressure difference across the membrane, AC = salt concentration difference, A = membrane surface £irea, and
5 = effective membrane thickness (pore size).
Advantages to using a RO membrane process for wastewater treatment are
1. removal of all ions and most dissolved non-ions,
2. responds well to changes in demand (flow and TDS level), and
3. ability to produce ultrapure water.
Disadvantages of an RO system include
1. a high capital cost,
2. the high level of pretreatment required for the RO make-up water (the water that
enters the RO system), and
3. the membranes may foul easily (Montgomery, 1985; Clifford, 1986; Aurich, 1995;
Chang, 1996; and Dow, 1997).
25
2.5 Additional Treatment Options
Other treatment possibiUties considered include evaporation, demineralization,
ion exchange, and contact fihers. Those processes are described briefly in this section,
but were not considered in the final proposal due to either their low ability to
effectively remove fluoride or for economic considerations. These systems are
presented here as alternative treatment processes.
2.5.1 Evaporation
Low-temperature evaporation systems use temperatures of 150 to 160°F to
produce steam; thereby, heating water under a vacuum system (Lindsey, 1993). A
distillation process follows the heating stage leaving behind the unwanted chemicals.
Distillation is an expensive, high energy, consuming process that uses hquid-gas
separation to remove a contaminant fi'om solution (Li, 1992). This is a highly effective
process with proven consistent productive results for the removal of highly
concentrated wastes in a variety of wastewater streams (regardless of chemical
concentration). However, the capital cost is high ($140,000) and requires a high
energy consumption ($20/1000 gallons of treated water) (Lindsey, 1993). Despite its
high efficiency rating, the rate of return on this system («6.9 years) makes it an
unlikely candidate for wastewater treatment in industrial settings where cost reduction
is a bottom line.
26
2.5.2 Demineraliziitinn
Demineralization is a process achieved by ion exchange, membrane processes,
or fi-eezing for an end product of water with no dissolved salts (CUfford, 1986; Li,
1992). Ion exchange methods or absorption (contact beds) are effective removal
techniques for water containing fluoride that needs to have a high removal efficiency,
which leaves a low concentration product (Chfford, 1986). Ion exchange is a selective
removal process for cation and anions whereby wastewater passes through contact
beds containing exchange resins. Most resms can be regenerated and can withstand
thousands of cycles before replacement is required (Clifford, 1986). The beds can be
regenerated when exchange capacities have been met. Ion exchange is used to obtain
higher levels of treatment than obtained by precipitation alone. The downside of ion
exchange methods are the high energy costs associated with the process and the
regeneration of the contact media.
2.5.3 Filtration
In filtration, particles are removed fi-om water as the water percolates through
granular media (Quasim, 1998). This media can consist of many different types of
materials (heterogeneous) or of one material (homogeneous using sand, anthrache,
lecithin, etc.) (Stuart, 1937; Bassett, 1973). The following are examples of absorption
removal techniques used for water containing fluoride:
1. Contact fihers 15" high and filled whh river sand passing a 60 mesh screen
27
mixed with two percent aluminum. Fluoride is removed by the absorptive qualhies of
the sand. The problem with this technique is that the regeneration rate of the sand
leads to a low efficiency rate for the filters and to an eventual solid waste disposal
factor (Stuart, 1937).
2. The second alternative is filter columns filled with clay which has the
capacity to remove fluorides fi-om a level of 1.1 mg/L to 0.46 mg/L The clay had a
high anion demand that correlated to the high cation exchange capacity of the fluorides
for an immediate removal by ion exchange. The downside of this technique is the high
drop in removal in a relatively short period of time. Once the clay meets its holding
capacity for cations, it stops removing them fi^om the wastewater. Regeneration and
sohd waste disposal are two additional drawbacks (Bassett, 1973).
2.6 Disposal of Concentrated Wastewater:
Regardless of the process selected for the treatment process, the by-products
of each process (whether it is coagulation, filter backwash, regeneration wastewaters,
or precipitation) are a small portion of the original wastewater streams, usually around
3 to 10 percent (Quasim, 1998). The disposal of the concentrated wastewater in the
discharge stream falls under the jurisdiction of the Federal Water Pollution Control
Act Amendments of 1972 and EPA fluoride discharge regulations, which specify a 32
mg/L daily maximum and a 17.4 mg/L daily average over a 30-day period for fluoride
m the discharge stream (48 Fed. Reg. 15394, April 1, 1983) (Koblyinski, 1997).
28
The most attractive form of disposal is to the local sanitary sewer system as
long as the concentration of the wastewater does not adversely affect the operations at
the local wastewater municipality (Qasrni, 1998). It has been found at newer
semiconductor plants (buih within the past 5 to 7 years) that the local wastewater
plants have increased treatment efficiency after installation of the recycling system at
the semiconductor facility (Williams, 1997).
Deep well disposal is an additional disposal option. This is a far less attractive
method of disposal and is considerably higher in cost. This disposal method is
regulated by local environmental regulations subject to geological and groundwater
studies. There are several other options that can be used as disposal techniques for
wastewater residual products (Quasim, 1998). Each facility should choose the method
of disposal that best suites their needs.
2.7 Examples of Water Management Programs
Several programs have been launched in the industrial arena to conserve water
and financial resources. The subsequent two sub-sections are brief descriptions of
case studies, their water quality return, and the economic savings resulting fi-om
implementing these treatment plans as a part as water management program. Each
case study is unique but points to the obvious advantages of a WMP. They have been
included in this Uterature review as additional illustrations of the effectiveness of a
WMP on industry.
29
2.7.1 Case Study One
In 1995, a printed circuit board manufacturer began a fiiU scale, reverse
osmosis based WMP. An average return with the WMP was one million gallons of
water each month. This was approximately 65 percent of the facihty's daily water
requirement. This water was recycled at 13 to 30 micro-ohm per cm quahty.^rx Oil Ui ior^ -Ko
/Production at the facility increased 52 percent with the increased quahty of the .
Figure 4.1: Case Study Logic Flow Chart (Source: Montgomery, 1995)
43
Included in the identification of the case study problem was a definition of the
experimental procedures, level of technology in use and available, and the
experimental parameters. Each of these parameters are important to better understand
and evaluate the identified problem, after which, the problem can be analyzed and a
solution formulated.
4.1.1 Experimental Procedures
Industrial wastewaters are constantly changing in terms of chemical waste
concentrations. Treatability studies were conducted beginning with laboratory tests
for the purpose of translating experimental data into design and operational
parameters. The first step was sample coUection foUowed by laboratory analysis. A
total of three sample collection techniques were used. Finally, statistical and
economical analyses were conducted to determine the process with the best potential
for energy efficiency and revenue return.
4.1.2 Level of Technology
In the case study, the industrial plant is operating 24-hours a day, seven-days a
week. The waste produced is segregated into an industrial waste stream and an acid
waste stream. The industrial waste stream was the primary focus. This waste is
collected from an accessible pump/lift station where aU industrial waste lines termmate.
The combined wastewaters are then pumped out for treatment and final discharge
from the plant. It was found not to be economical or time permissible to segregate the
44
waste Unes contributing to this hft station for sample collection, based on man hours
and the potential hazard firom breaking lines. Therefore, samples were collected from
the sump at the wastewater lift station.
4.1.3 Experimental Parameters
It was estimated that each plant operation contributing to the final industrial
waste stream (primarily those within the wafer fab) would dump waste at least once
every four hours. The hft station, where the waste streams came together and samples
were coUected, pumped continuously at a rate based on the volume of the wastewater
within the tank. Therefore, h was determined that the most accurate form of sample
coUection would be a continuous sample coUection operation. This, however, is not a
possible option as the equipment and man-hours were not available for this project.
Therefore, with each of the following listed factors considered, the first set of samples
were coUected every hour during a 24-hour samphng period.
1. Plant operations contributmg to the final industrial waste stream dump at least
once every four hours. It is desired that each constituent and its concentration in
the waste stream be sampled at least one time during the mterval.
2. Sample coUection required approximately 20 minutes.
3. Lab analysis for each sample was estimated at one hour (for analyzing samples in
triplicate). Metals, salts, TS, and TDS were all analyzed at the parts-per-nuUion
level. Conductivity and pH were also included in the lab tests.
45
4.2 Analysis of the Case Study Problem
The analysis of the-how-to-approach the water management program began
with meetmg three objectives: a water inventory, sample coUection, and an inventory
of the contaminants in each of the samples. Once these objectives had been met, a
solution could be formulated to design the WMP.
4.2.1 Water Inventory
The water inventory was the fhst step in estabUshing a need for recycling water
within the faciUty. Therefore, a water inventory was made of the high water demand
operations whhin the facUity. Those areas are Usted on Table 4.1.
Table 4.1: Approximate Water Inventory
Process Scmbbers
Coolant Towers (North) Coolant Towers (South)
DI Plant
# Units 9 11 6 1
Demand per Unit 10-15gpd 5500 gpd 2700 gpd 150 gpd
Sut3total 135
60500 16200 150
(Source: Hector, 1997)
The operations listed are not aU inclusive of every process or operation whhin
the plant that requhes water. However, they are major supply and demand areas. The
water demand for these processes alone comes to almost 77,000 gpd. This is the
amount of water purchased and discharged on a daUy basis. Thus, when the water
demand of these operations are combined, they create a major contribution to the
recycling process.
46
4.2.2 Sample Collection
The first step was to construct a plan to determine which areas of the facUity
samples would be coUected. Safety requhements and Standard Methods (method
300.0), techniques were foUowed for coUectmg the samples. The mdustrial
wastewater whhin the facUity faUs into one of three categories: acid wastewater,
industrial wastewater, or water from the DI plant.
To best represent the water being evaluated and sufficient quantities collected
for analysis; aU major contributing water and wastewater supply lines, the two main
wastewater Uft stations, and several lines fi'om the DI plant were tested to determine
the quahty of the avaUable water supply. Initial samples were coUected from each
main industrial and acid wastewater stations and lines on random days and times over
the course of six months (see Figure 4.2). These initial samples were used to create
an overview and baseline of the contaminants present in the industrial and acid
wastewater supply. The wastewater within this faciUty feU into ehher the acid or
industrial waste category.
Both the EPA and Standard Methods procedures recommend the use of ehher
polyethylene or glass containers for coUection and transport of materials that could
contam hazardous contents such as those in the wastewater streams presented in this
case study. Therefore, each sample was coUected, transported, and stored in
sterilized, polyethylene bottles (150 and 500 ml). In addition, the EPA guidelines
recommended that samples were not stored for more than 28-days (14-days for
Nhrate-N, non-chlorinated water) before lab analysis could be conducted. In
47
following these guidelines, tests were performed withm one week of theh coUection
date.
Since hazardous and/or corrosive materials were possible in either of waste
streams, a hazardous condhions plan was constructed by the she's safety engmeer.
The plan included the recommendation of using personal protective equipment (PPE):
safety glasses, face shields. Tan Tionic® gloves over gray 4H gloves, gray 4H apron,
and gray 4H sleeves. Any disposable material that contacted the waste material was
disposed of within 30-minutes of use. The foUowing is a Ust of the locations within
the semiconductor plant where samples were initially collected.
1. Incoming Tap Water 2. Industrial Waste Tank No. 2 3. RO Brine 4. RO Product 5. ROMake-Up 6. Recycled DI 7. Industrial Waste Lift Station 8. 125 mm Scrubber
9. 150 mm Scrubber 10. TLM Scrubber 11. Acid Waste Lift Station 12. Wet Processes 1, 2, and 3 13. Wet Processes 4 and 5 14. Lateral One 15. Clean Up Shop
Figure 4.2: Sampling Shes Whhin the Facility
On November 26, 1996, samples 1 through 10 were coUected and analyzed.
On December 10, 1996, samples 1 through 6 were collected and analyzed to confirm
the initial set of data. A more specific analysis was necessary to locate point source
pollution areas. Therefore, on June 15, 1997 samples 2, 5, 7, and 11-15 were
collected and analyzed agam. It was evident at this point that there would be ample
areas fi"om which wastewater could be recovered and coUected that was of treatable
quality within the facility to treat and recycle.
48
The wastewater in each line contributing to the lift stations flows under
gravitational forces. It was important to segregate mdividual Unes contributing to the
acid lift station, as the waste contained there was strong and highly concentrated. If
mformation could be obtained that would mdicate which Ime or Unes contributed a
higher concentration of waste to the final waste stream, then that would allow
rerouting of one or more of these Imes so that the remaining volume of wastewater
could be used for recycling purposes. There are four main lines that contribute to and
come together at the acid lift station. Those four acid wastewater lines are on Figure
4.2. shown below.
Lateral Line #1
i Wet
Processes 4 &5
i Wet
Processes 1,
i Acid Lift Station
2, &3 Ciean-Up
Shop
i
Figure 4.3: Acid Wastewater Flow Scheme
The combined flow rate out of the acid lift station is a small quantity compared
with the flow out of the industrial lift station. After laboratory analysis of the
segregated acid wastewater lines (Table 4.2), h was determined that the strength and
concentration of the acid wastewaters (in part due to the decrease in dUution resuhing
in an overaU smaller volume) would be too high to treat and recycle at an economic
rate.
The wastewater lines that contributed to the industrial hft station could not be
segregated as easily as the acid lines. They are more numerous in number and have
49
Table 4.2: Acid Wastewater Analysis
Sample
Acid Line 1
Acid Line 2
Acid Line 3
Acid Lift Station
Fluoride
ppm
213 900
63 9 465
Chloride
ppm
171 595
ND NO
Nitrate
ppm
ND ND
ND ND
Sulfate
ppm
14416 16517
6 15928 13180
Sodium
ppm
8776 15847
41038 40557
Potassium
ppm
8522 16684
37527 25034
Calcium
ppm
9101 9246
14579 18203
Iron
ppm
6148 3583
6454 5922
pH
1.20 1.10
1.20 1.60
many coimecting points. Therefore, they are not individuaUy represented, but are
represented as one wastewater stream that contributes to the final product at the point
of the mdustrial Uft station. It is not necessary to show the Unes contributing to the
mdustrial lift station, as the wastewaters are not as concentrated. The general flow of
the existing mdustrial and acid wastewater within the faciUty and the discharge from
the plant can be seen in the flow diagram (Figure 4.4).
Acid Lift Station
i-
Industrial Lift Statior
-i
i Addition of Caustic
Discharge to City Wastewater Treatment Plant
Figure 4.4: General Wastewater Flow Scheme
Three sample collection techniques were used to determine a comprehensive
analysis of the wastewater stream. Those three techniques were:
50
1 A 24-hour sampling period, in which samples were collected every hour (on the
hour) for 24-hours for a total of 24 samples. Grab samples were collected with a
hand-held pump inserted in the center of the industrial waste tank approximately
halfway from the water surface. This approach was a hit-or-miss technique and
had a low probability of returnmg reliable data. This sample collection set will be
referred to as data set A
2. In hopes of increasing accuracy, a second approach utilized a metered pump
secured over the center of the tank with the intake tubing anchored near the
bottom of the tank. Samples were continuously pumped out at an approximate
rate of 5 ml per minute and composite samples were collected every thirty minutes.
The coUection continued for 24-hours for a total of 48 samples. The reliability of
the data was expected to increase over that of the first technique due to the
continuous samphng approach even though analysis was not instantaneous. This
collection technique was used to gather data continuously over one randomly
selected day. This data set will be referred to as data set B.
3. The third set of samples were collected utilizing the second approach of
continuous samphng whh a metered pump but over seven consecutive days.
Samples were continuously pumped at an approximate rate of 5 ml/minute for each
24-hour period and then composite samples were collected at the end of each 24-
hour period. This technique gave a total yield of seven samples. These samples
were collected to examine the variance and central tendencies from a day-to-day
stand point. This data set will be referred to as data set C.
51
4.2.3 Inventory of Wastewater Contaminants
Metals and salts were the two primary categories of analysis. Metals in water
and wastewater range in their effects on treatment processes and the surrounding
environment from beneficial to dangerous. Dissolved inorganic contammants can be
amphoteric, neutral, cationic, or anionic forms of ions, atoms, or particles of any
element in the periodic table. The first step in testing each sample was an analysis for
fluorides, chlorides, nitrates, and sulfates using a Dionex Ion Chromatograph series
40001 with Al-450 software. Each sample was then tested for metals including Na, K,
Ca, Fe, Mg, Ni, Zn, and Sn (among other metal elements) usmg atomic mass
absorption methods.
To complete the lab analysis, each sample was tested for conductivhy, pH,
total dissolved sohds, and total solids. Each data set coUected from December 1996 to
October 1997, was analyzed by each of these methods. A comprehensive analysis of
all of these lab tests was used to obtain a wastewater quahty level of the industrial
waste. Qualitative characteristics were necessary to determine the most cost-effective
and cost-beneficial treatment process.
4.2.3.1 Sah Analysis
Total dissolved solids (TDS) is the sum of the dissolved salts in solution.
Therefore, the samples were first analyzed for salt contents. In this group, fluorides
created a specific problem in the treatment process. The most electronegative of aU
52
the elements, fluoride is almost always combined whh other elements when found in its
natural envhonment (Onuoha, 1983).
The POTWs that supply the water used in this case study produce a water
supply that meets the U.S. PubUc Health Source Drinking Water Standards mandatory
limit for aUowable fluoride. This pretreatment goal is to reach this level of fluoride
concentration or the maximum permissible concentration (as noted by the membrane
manufacturer) in the mdustrial waste before sending h to the RO system. Salt levels
for each of the sample sets coUected are found on Tables 4.3, 4.4, and 4.5.
4.2.3.2 Metal Analysis
Hardness of the water is the sum of the dissolved iron, dissolved manganese,
calcium, and magnesium. Sodium, potassium, calcium, and hon were the primary
dissolved inorganics of concern in the industrial wastewater. In this case, hon was the
only constituent that peaked over the city concentration regulations. The hon level
was close to the regulation level and does not pose serious treatment problems as does
fluoride. The metal analysis for each of the samples collected can be found on the
following pages on Tables 4.6, 4.7, and 4.8.
4.2.3.3 Addhional Analysis
Each of the samples were then analyzed for pH, conductance, TDS, and TSS.
The coUection of these chemical and physical parameters were used to better
determine the appropriate treatment process or series of processes.
Where: ai=[(nSxi'-(Ixi) 0/(n*(n-l ))]^. 5 N=(t')*(6^/8^) ReliabUity ( r ) > 50% 6= standard deviation e= tolerance of utUized equipment t= student's distribution and rehabihty table N= Number of tests x= percentage of rehabihty (Cohen, 1979)
The data on Table 4.11 can be used to determme the most rehable sample
coUection method. These values indicate the samples' reliabUity whhm each
percentage range. The lower the value on Table 4.11 for each constituent, the smaller
the distance that value is fi-om the mean for that contaminant (within that particular
percentage of rehabUity). As the reliability increases (85 percent^90 percent->95
percent), the values on the table decrease. This is because the distance fi'om the mean
decreases (Gilman, 1997). The smaller the value calculated for each percentage range,
63
the higher the reliabUity. Therefore, when makmg comparisons of the reliability
between the three data sets (A, B, and C), the smallest value for each percentage range
exhibhs the best rehability. This was the procedure followed to identify the most
reliable sample coUection set.
From Table 4.11, data set C has the highest reliabUity. Therefore,
the sample collection information fi'om this will be used m the subsequent
statistical analysis step.
Table 4.11: Rehabihty Levels of Data Sets A, B, and C
Data Set (C): 24-Hr sampling, 7 day period (9/26/97-10/3/97)
Reliability (0.85)
Reliability (0.90)
Reliability (0.95)
Reliability (0.99)
Metals (ppm)
Na
3.70
5.70
6.70
7.70
K
0.00
2.00
3.00
400
Ca
0.21
2.21
3.21
4.21
Fe
1.47E-05
2.00
3.00
4.00
Salts (ppm)
Fluoride
8.27
10.27
11.27
12.27
Chloride
11.17
13.17
1417
15.17
Nitrate
0.11
2.11
3.11
411
Sulfate
7.21
9.21
10.21
11.21
Phosphate
42.64
44.64
45.64
46 64
Data Set (A): 24-Hr sampling (0700 8/8/97 • 0600 8/9/97)
Reliability (0.85)
Reliability (0.90)
Reliability (0.95)
Reliability (0.99)
M^als (ppm)
Na
6.62
8.62
9.62
10.62
K
0.00
2.00
3.00
4.00
Ca
0.08
2.08
3.08
4.08
Fe
0.06
2.06
3.06
4.06
Salts (ppm)
Fluoride
18.36
20.36
21.36
22.36
Chloride
144.76
146.76
147.76
148.76
Nitrate
0.04
2.04
3.04
4.04
Sulfate
69.50
71.50
72.50
73.50
Phosphate
Data Set (B): 24-Hr sampling (0700 9/12/97 - 0650 9/13/97)
Reliability (0.85)
Reliability (0.90)
Reliability (0.95)
Reliability (0.99)
Metals (ppm)
Na
43.55
46.55
46.55
47.55
K
0.66
2.66
3.66
4.66
Ca
0.02
2.02
3.02
4.02
Fe
2.53
4.53
5.53
6.53
Salts (ppm)
Fluoride
2.09
4.09
5.09
6.09
Chloride
113.84
115.84
116.84
117.84
Nitrate
0.15
2.15
3.15
415
Sulfate
63.42
65 42
66 42
67.42
Phosphate
2219.52
2221.52
222252
2223.52
64
4.3.1.2. Determinmg the Quahty of Sample Sets
The second stage of the statistical analysis was determmmg the quahty of the
selected sample set C. This was done to ensure that the set feU whhin acceptable
quahty control (QC) hmits. Just because the sample set was selected as the most
rehable of the three sets, hs rehabUity did not indicate that h was also whhin QC Ihnits.
If all three of the samples had been outside the QC hmits, then the set selected
as most rehable of the three stUl would not have been very dependable. Therefore,
conventional quahty control charts were prepared for each of the consthuents in
sample set "C." The data pomts used to construct the QC charts can be found on
Table 4.12.
Table 4.12: Conventional Quality Control Data for Sample Set C
standard Deviation
From the mean
Mean
Standard Deviation
(+) 1 (good)
(+) 2(accept.)
(+) 3 (poor)
(-) 1 (good)
(-) 2 (accept.)
(-) 3 (poor)
Metals
Na
40.17
11.10
51.30
62.40
73.50
29.10
18.00
6.90
(ppm)
K
0.76
0.34
1.10
1.44
1.79
0.42
0.08
-0.27
Ca
5.33
2.67
7.97
10.60
13.31
2.66
0.00
-2.67
Fe
0.03
0.02
0.05
0.07
0.09
0.00
-0.02
-0.04
Salts (ppm)
Fluoride
46.64
16.60
63.20
79.80
96.40
30.00
13.40
-3.20
Chloride
105.46
19.30
124.80
144.10
163.40
86.20
66.90
47.60
Nitrate
0.73
1.93
2.66
4.59
6.52
-1.20
-1.93
-2.66
Sulfate
107.47
15.50
123.00
138.50
154.00
92.00
76.50
61.00
Phosphate
127.73
37.70
165.40
203.10
241.00
90.00
52.30
1460
** Samples from 24-Hr sampling, 7 day period (9/26/97 -10/3/97)
The QC charts were a usefiil tool for determining whether the coUected data
was legitimate or simply an experimental error for quanthatively hmited data. Outliers
were identified and each consthuent withm each sample was classified on a qualitative
level. The QC charts are based on the variance of the consthuent from hs mean value.
This distance fi-om the mean (based on standard deviations) determined a sample's
65
acceptance or rejection fi-om the coUected set. Table 4.12 is an outlme of the
quahtative parameters assigned to each constituent and was used as the bases for the
QC charts (Figures 4.5, 4.6, and 4.7).
Sample set A was determined to be an acceptable data set based on the
mformation fi-om the QC charts and the hmits as set on Table 4.12. Each of the charts
showed that the samples were within acceptable quahty levels of the samphng
techniques. Therefore, the information fi'om data set A was used m the third stage of
The data obtamed fi-om the present worth cost wiU be used later in this section
to relate the benefit-cost ratio. The present worth cost is calculated by Equation 4-4.
It involves the use of a discount rate for each of the processes. The discount rate is a
constant whh a time value for capital and the operational and maintenance cost of each
of the alternatives. This rate is commonly used by the federal government, but can be
interchanged with the interest rate, commonly used by industry. The interest rate used
by most industries and determined by the capital market, includes the tune value of
money and a risk factor (James and Lee, 1971). For this study, the interest rate was
assessed at 7,10, 12, 15, 20 and 30 percent to ensure economic attractiveness of the
WMP to an industry. The results are on Table 4.18.
Table 4.18: Summary of Present-Worth at Various Interest Rates
Present Worth ($)
Interest Rate (%) Lime
Activated Alumina Electrocoagulation
(year =1997)
7 269,000
1,198,000 1,052,000
10 273,000
1,077,000 864,000
12 279,000
1,029,000 778,000
15 291,000 987,000 684,000
20 319,000 970,000 584,000
30 392,000
1,040,000 480,000
Choosing a low interest rate favors projects with high capital cost, such as the
EC system. High interest rates favor low caphal costs and operation and mamtenance
(as whh lime and activated alumina). When assessing a project whh high caphal cost
(the marginal return in an alternative use project), the opportunity cost of the project
should be taken into consideration (James and Lee, 1971). Therefore, the benefits to
77
processes downstream from the project (such as the RO unit) are added to the cost-
benefit ratio analysis.
Efficiency of one pass through an RO unit can be increased from 65 percent up
to 95 percent through the use of the EC process. If the efficiency of the unit is
increased 30 percent, that is 30 percent less work that the RO unit is required to
perform. That same system would then have a probable life expectancy mcreased by
up to 30 percent. Basically, the same system would be able to operate 30 percent
longer at the same O&M cost after hs efficiency has been increased by 30 percent.
For a 300-gallon per minute system, the savings are $67,600 each year. When O&M
for an EC system are calculated at $63,072 each year, this leaves an addhional benefit
of $4,428 each year. This is in addhion to the $409,968 that is saved each year from
purchase and disposal costs by each of the three proposed systems when water does
not have to be replaced (at $2.60 per 1000 gallons) after each pass through the facility.
4.3.3.2 Cost/Benefit Ratio
The benefit-cost analysis is a ratio of the present-worth benefit to the
present-worth cost. When using the present worth method to make comparisons
between projects the rule is to select the project whh the largest present worth:
I Benefits - Cost of Life of Project = B/C. Equation (4-4)
This is true assuming that all present worth ratios are calculated at the same
point in time, use the same discount rate, and are analyzed for the same period of time.
78
This follows the four rules outlmed for cost-benefit analysis (see Section 3.3.3). Cost-
benefit comparisons were made between the three processes (Table 4.19). When
making a C/B ratio, the lowest ratio is the most desirable process.
Tables 4.18 and 4.19 should be constructed for each case study to make final
economic considerations. From the data on these tables, EC and lime present the most
economicaUy desirable solutions in the design of a WMP At a 30 percent interest
rate, the electrocoagulation and lime processes are cost competitive. However, as the
interest rate decreases the lime process becomes the more beneficial process.
Table 4.19: Summary of Cost/Benefit Ratio at Various Interest Rates
Interest Rate (%) Lime
Activated Alumina Electrocoagulation
7 0.66 2.92 2.20
10 0.67 2.63 1.81
12 0.68 2.51 1.63
15 0.71 2.41 1.43
20 0.78 2.37 1.22
30 0.96 2.54 1.01
4.4 Case Study Summary
Once treatment options were narrowed down to five main processes
(precipitation, electrocoagulation, absorption/ion exchange, evaporation, and
membrane separation), basic considerations of the design were accounted for in the
treatment system. Characteristics that were taken into consideration included the
quality of the raw water supply (in this case the industrial wastewater), she conditions
(indoor space, available land addhions to the existing building, topographic, etc.),
plant economics, and regulatory constraints on the discharge of wastewater.
79
Technological feasibhity was the first consideration but the economic evaluation
should be the fmal deciding factor in process selection.
4.5 Case Study Recommendations
Ultimately, process selection should be a decision made by the facility
implementing the program. The purpose of the WMP in this case study was to narrow
down and recommend treatment options. From the economic evaluation and the
technical analysis, hme preciphation and electrocoagulation appear to be the options
best suited for this situation.
The final decision on whether or not to implement the program rests whh the
facility. The technical analysis and economic evaluation presented in this research
should prove to be of merit in providing evidence to support the implementation of a
water management program.
80
CHAPTER V
CONCLUSIONS AND RECOMMENDATIONS
5.1 Conclusions
The results of this study showed that:
1. Recycled wastewater can increase the productivity of a facility by increasing the
quahty of the mitial water supply. This is m addhion to the monetary incentives
provided by the hnplementation of a water management program. This is an
important consideration in the highly competitive fields of electronics and
semiconductors.
2. The implementation of a water management program is a technically and
economically feasible project for semiconductor industries with high water
consumption and wastewater discharge.
3. The general procedures developed in this study are non-specific and have
apphcation beyond those of the semiconductor industry. Recycling projects should
be designed to first meet technical feasibhity and then each process should be
evaluated as to economic merit.
On the basis of these three conclusions, water management programs should be
pursued by industries interested in saving both the tangible and intangible factors of
water conservation, reuse and recyclmg.
81
5.2 Recommendations
DetaUed research is required in the design of a water management program in
order to meet the specific needs of an hidustrial facUity. The general procedures
outlined in this study have been shown to be effective both technically and
economicaUy effective on a bench scale level. The bench-scale analysis in the case
study should be followed by pilot studies, as in any fuU scale study, before
implementation of the selected alternative. The ultimate treatment selection is up to
the industrial plant implementing the program.
Additional studies should be apphed to the economic incentives added to the
program through:
1. The increased operating efficiencies of a facility due to the increased quality of the
recycled water when replacing the initial feed water supply (tap water), and
2. Economic benefits of harvesting and seUing the metals and salts recovered through
the treatment processes.
82
REFERENCES
Alvarez-Lara, Fernando, Facihties LMOS Engineering Group. Personal Communication. (1997).
Amjad, Zahid. Reverse Osmosis. Membrane Technology. Water Chemistry and Industrial AppHcations. New York: Van Nostrand Reinhold. 1993.
Aurich, Christoph W. Wastewater treatment-Choosing the Right Membrane System. Journal of SDC. Volume 111. June 1995.
Bassett, Randy Lynn, A Geochemical Investigation of Sihca and Fluoride in the Unsaturated Zone of a Semi-Arid Environment. Master's Thesis, Texas Tech University, 1973.
Bulusu, K.R., and Pathak, B.N., Water Defluoridation whh Activated Alumina. EE2. April 1980.
Chang, Dong-Jang, and Hwang, Shyh-Jye, Removal of Metal Ions from Liquid Solutions by Crossflow MicrofUtration. Separation Science and Technology Marcel-Dekker, Inc. 1996.
Clifford, D., Suresh, S., and Sorg, T. Removing Dissolved Inorganic Contaminants form Water. Envhonmental Science Technology. Vol. 20, No. 11, 1986.
Cohen, Jacob, Statistical Characterization of Construction Materials. Moscow: Construction Pubhshing House, 1979.
DiUon, C.P. Materials Selection for the Chemical Process Industries. New York: McGraw-HiU, 1992.
Gaeta, S.N., The Industrial Development of Polymeric Membranes and Membrane Modules for Reverse Osmosis and Ultrafihration. Membrane Technology: Apphcations to Industrial Wastewater Treatment. Netheriands: Kluwer Academic Pubhshers. 1995.
Oilman, Alex, Associate Professor, Engineering Technology, Texas Tech University. Personal Communication. (1997).
Global Water Industries. (1997). Custom-Engineered Solutions for Water Cleanup and Resource Recovery for Industry and Municipalities [on line]. Available: http:/www.globalwaterindustries.com/home2 shtml
Hector, Jerald Facihties Employee, Texas Instruments, LMOS. Personal Communication. (1997).
James, Douglas, and Lee, Robert. Water Resources and Environmental Engineering. New York: McGraw-HiU, 1971.
Johnson, Richard A., MiUer and Freund's Probabihty and Statistics for Engineers. New Jersey: Prentice Hall. 1994.
Maier, F.J. Methods of Removing Fluorides from Water. American Journal of Public Health. Volume 37, December 1947.
Mann, PhD Chemical Engineering Department, Texas Tech University. Personal Communication. (1997).
McKee, R.H., and Johnston, W.S., Removal of Fluorides from Drinking Water, Journal of Industrial and Engineering Chemistry. 26 (August 1934): 849.
Memtec (1997). [Laboratory bench scale, test results of lime preciphation]. UnpubUshed raw data.
Mollhagen, Tony and Harkins, Victoria, Water and Wastewater Analysis: CE 5374. [Unpubhshed manual]. (Spring 1995).
Montgomery, James M., Water Treatment: Principles and Design. New York: Wiley-Interscience Pubhcations. 1985.
Onuoha, Ukiwo Obasi, "Evaluation of an Activated Alumina Sorption System for Removal of Fluoride from Water." Master's Thesis, Texas Tech University, 1983.
Qasim, Syed R., Notes from Wastewater Treatment Plants Planning. Design, and Operation. Unpubhshed notes (expected out 1998).
SavineUi, Emiho A., and Black, A.P. Defluoridation of Water with Activated Alumma. Journal of American Water Works Association. January 1958.
Sorg, T.J., Treatment Technology to Meet the Interim Primary Drinking Water Regulations for Inorganics, Journal of American Water Works Association. 70 (February 1978): 105-112.
Stuart, T. H., "The Removal of Fluorides from West Texas Water." Master's Thesis, Texas Tech University, 1937.
Tock, Richard Ph.D. Chairman of Chemical Engineering Department, Texas Tech University. Personal Communication. (1997).
United States Fiher Corporation, [unpubhshed busmess material]. 1997.
Varuntanya, C. Peter, and Shafer, Dennis, Lime Preciphation Insufficient for Removing Fluoride. Industrial Wastewater, pp. 32-35. January/February 1994.