United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-84-018 Mar. 1984
Project Summary
A Mobile Drinking Water
Treatment Research Facility for
Inorganic Contaminants
Removal: Design,
Construction, and Operation
Dennis Clifford and Maheyar Bilimoria
A mobile inorganics removal research
facility consisting of a pilot plant and
analytical laboratory was designed and
constructed and has been operated for
3 years. Ion exchange, activated alumina
adsorption, reverse osmosis, and elec-
trodialysis have been studied in this
transportable facility for the removal of
fluoride, nitrate, and chromium. Plans
call for the study of arsenic and
selenium and any other inorganic con-
taminants of interest. To date, the
facility has performed very well and
much valuable pilot-scale data have
been obtained.
This Project Summary was developed
by EPA's Municipal Environmental
Research Laboratory, Cincinnati. OH,
to announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
An estimated several thousand public
water supplies in small communities in
the United States contain fluoride,
nitrate, arsenic, selenium, radium,
barium, or chromium in concentrations
exceeding the maximum contaminant
limits (MCL's) established in the National
Interim Primary Drinking Water Regula-
tions. Previous experience and research
indicate that these primarily ionic con-
taminants can be removed by advanced
water treatment processes such as
packed beds of activated alumina, ion-
exchange resins, or by separation using
reverse osmosis or electrodialysis. But
reliable design criteria and economical
operating procedures are not available for
the selection, cost-effective application,
and safe operation of these processes.
Such is especially true for small com-
munity water supply treatment, where
single contaminant removal is desired in
waters containing more than 1000 mg/L
of total dissolved solids.
To provide help for small communities,
a long-term, EPA-funded project has
been undertaken to evaluate the single
contaminant removal processes (activated
alumina and ion exchange) versus the
desalting processes (electrodialysis with
reversal and reverse osmosis). To accom-
plish this project, contaminated water
sources in a series of small U.S. com-
munities are being studied with the use of
a mobile drinking water treatment
research facility.
This report describes the field research
capabilities of the 3.2- by 12.5-m (10-by
40-ft) transportable research facility and
summarizes its design, construction, and
operation. The mobile facility contains a
complete analytical laboratory and an 8-
L/min (2 gpm) pilot plant with inter-
connected reverse osmosis, ion exchange,
activated alumina, and electrodialysis
units. The treatment technologies appli-
cable to a given contaminant removal
problem are operated separately over a
period of several months. The resulting
-------
performance data for all processes are
then compared on technical and economic
bases, and appropriate general recom-
mendations are made for that type of
contaminant removal problem. To date,
fluoride, nitrate, and chromium removal
have been studied in Taylor, Texas;
Glendale, Arizona; and Scottsdale,
Arizona, respectively. The 1982 replace-
ment cost of the facility is approximately
$200,000.
Inorganic Contaminant
Removal Processes
Five of the inorganic contaminants
listed in the EPA National Interim Primary
Drinking Water Regulations are anions or
generally occur as anions in the waters
where they are found. These contaminants
are fluoride, nitrate, arsenic, selenium,
and chromium. Details of their speciation
as a function of pH appear in Table 1
along with their MCL's.
Two distinct types of processes may be
considered for removing these ions from
drinking water: single contaminant
removal processes (e.g., alumina adsorp-
tion and ion exchange) and desalting
processes (e.g., electrodialysis and
reverse osmosis). Single contaminant
removal generally costs less, but desalting
yields a softer, more palatable, and less
corrosive water for distribution. Each
contaminant removal application will be
different because the background raw
waters have different total dissolved
solids levels and chemical compositions.
Both desalting and single contaminant
removal processes should therefore be
available when doing contaminant re-
moval research.
The literature indicates that fluoride,
arsenite, arsenate, selenate, and selenite
can be removed from low or high total
dissolved solids water supplies by
precipitation or coprecipitation with lime,
alum, or iron salts. Unfortunately,
precipitation processes are not readily
adaptable to small water systems or
individual wells that must operate on
demand. The negative features of coagula-
tion-precipitation processes in these
applications include the need for sludge
collection, dewatering, and disposal and
long start-up and shut-down periods.
Fluoride and arsenic have been removed
from water supplies using packed beds of
bone char or tricalcium phosphate. But
these media tend to degrade by attrition
and continuously lose capacity after
successive regenerations. Packed beds of
activated alumina and anion exchange
resins have been chosen for this single
contaminant removal research. Such
Table 1. Potential Anionic Contaminants for Water Supplies
Contaminant MCL. mg/L Common Form in Ground Water (pH=6-9)
Fluoride
Nitrate (as N)
Arsenic
Selenium
Chromium
1.6"
10.
0.05
0.01
0.05
r
NOs
H2 AsO*. HAsO*. HAsOi. AsO2~
SeOf, SeO3~, HSeO3~
CrOi
"Annual average of maximum daily air temperature = 70.7 - 79.2°F.
columns can be operated on demand and
are generally free from the disadvantages
of precipitation processes and packed-
bed processes using bone char and
tricalcium phosphate.
For desalting small flows of brackish
water, electrodialysis and reverse osmosis
are the methods of choice. Electrodialysis
removes ions from water; reverse osmosis
removes water from ions. In both pro-
cesses, low-molecular-weight (<200),
uncharged species tend to pass through
into the product water. But properly
operated reverse osmosis units will
remove bacteria, viruses, and silica,
whereas electrodialysis units will not.
Finally, these desalting processes may be
expected to do a fair-to-good job of
removing the inorganic contaminants of
interest even in high (>1000 ppm) total
dissolved solids waters. The single
contaminant removal and desalting
processes to remove fluoride, nitrate, and
the various forms of arsenic and selenium
can be compared in Table 2. Entries in the
table are primarily based on theory and
bench-scale research findings, except for
fluoride removal, which has been practiced
on a large scale.
Mobile Research Concept
The reusable pilot-plant concept was
thought to be particularly applicable to
the diverse inorganic contamination
problems in small communities. In early
1979, work was started in earnest on
an EPA-funded research project at the
University of Houston (UH) to design,
construct, and operate a transportable,
reusable, pilot-plant facility for inorganic
contaminant removal. This is the design
and construction report for the transport-
able pilot plant, which was completed in
April 1980. As of December 1982, it has
been operated in Taylor, Texas, for
fluoride removal; in Glendale, Arizona,
for nitrate removal; and in Scottsdale,
Arizona, for chromate removal. Experience
from the first three moves has shown that ^
the facility is readily transportable and f|
reusable. The pilot-scale treatment
systems have operated very well, and
much valuable pilot-scale data have been
obtained.
Table 2. Potential* for Contaminant Removal by Various Treatment Processes
Packed Beds
Reverse Osmosis
Electrodialysis
Contaminant
Fluoride
Nitrate
Arsenic (III)
Arsenic (V)
Selenium (IV)
Selenium (VI)
Chromium (VI)
Activated
Alumina
pH 5.5 - 7.5
G+
P
F/P
G
G
F
F
Strong
Base
Resins
pH5-9
P
F
P'
P/G2
F3
G4
G
Cellulose Acetate or
Aromatic Polyamides
pH6-8
G
F/G
P/F
G
G
G
G
pH6-8
G
F/G
P/F
G
G
G
G
"Potentials are based on published experimental results except where noted:
'//a AsOs, the uncharged species, predominates at pH's below 9.2
2Poor at pH's below 7 (Hz AsOi). good at pH's above 7 (HAsOt°)
3Estimated, based on the elution of selenite in ion chromatography
'Estimated, based on the elution of selenate in ion chromatography
*G = Good F = Fair P = Poor: These are relative rankings for the process in question. For
electrodialysis, and reverse osmosis at 50% to 8O% recovery:
Good means greater than 80% removal in typical ground water.
Fair means 40% to 80% removal.
Poor means less than 40% removal.
For packed bed processes:
Good means that the ion is highly preferred relative to Cr,
Fair means that the ion is a preferred ion relative to Cr,
Poor means that the ion is not a preferred ion relative to Cr. .
-------
The mobile water treatment pilot-plant
system (Figure 1) consists of a 3.2 x 12.5-
m (10- x 40-ft) research trailer, a pick-up
truck containing a 1160-L (300-gal)
wastewater tank, and a 9.7-m (31-ft)
travel trailer. The research trailer (Figure
2) is transported between sites by a
professional tractor-trailer driver, and the
travel trailer is pulled by the pick-up truck
driven by the field researcher. While on
location in the field, connections are
made to the research trailer to supply raw
water (an average of 8 L/min), electrical
power (100A, 220V), and telephone
service. Unused treated waters and
nontoxic wastewaters are disposed of
Field Researcher
Living Quarters
Located in Nearby Trailer Park
31' Travel Trailer
Wastewater^
Telephone —
Flaw Water —
220V. 100 A-
300 Gallon
Wastewater
Tank — *•
o
)
/ /
1 Ton Pick-up Truck
Water Treatment Pilot Plant
RO. IX. EDFt, AAI. Pumps, Control
Panel. Ghent Storage and Prep..
Water & Wastewater Tanks,
Pretreatment Sys.. Workshop,
Tools, Spare Parts, Safety
Shower, Eye Wash, Fire Ext.
Office
and
Analytical Lab
pH. umho, TDS, JTU
SDI.r.NOl.Cr
TH. Alk, SOf, SiOi
Figure 1.
10x40'
Research Trailer
Mobile research concept including transportable pilot plant/laboratory, travel
trailer, pickup truck.
Figure 2. UH/EPA mobile drinking water treatment research facility shown hooked up to
contract hauler's tractor leaving the University of Houston on its way to Taylor, Texas.
through discharge lines to a nearby
sewer or by surface spreading (grass
watering). Toxic wastewaters (e.g.,
concentrated As or Se solutions) will be
pumped into an 1160-L (300-gal) tank in
the pick-up truck and transported to an
ultimate disposal site.
During the 3- to 12-month period at a
given field location, the field researcher
lives in the travel trailer. The latter is
generally located in a nearby trailer park
where complete utility hookups are
available.
Water treatment process research and
water analyses are both done in the large
research trailer, which is divided into two
sections (Figure 1). The rear two-thirds of
the space is devoted to the pi lot plant, and
the front third contains the analytical
laboratory and office. Ideally, the field
researcher is an environmental or
chemical engineer with analytical chem-
istry skills. He or she does both the pilot-
plant experiments and the water analyses
or supervises the water analyses. Other
personnel involved in the research
include the principal investigator (PI),
environmental engineering graduate
students, a part-time analytical chemist,
and outside contractors. The PI supervises
the design and execution of all experi-
ments by phone, letter, and site visits.
Graduate students are occasionally sent
to the field locations to assist the field
researcher for periods of 2 to 6 weeks.
Part-time chemists who are local residents
are used to assist in the analytical work
whenever possible. Finally, outside
contractors are hired to move the trailer,
install the electrical power, and maintain
the instruments and control systems.
Selecting Field Locations for
Study
Small community water supplies to be
evaluated for contaminant removal
studies are selected by the PI on the basis
of:
(1) Severity of the inorganic contami-
nant levels and demonstrated health
effects.
(2) Usefulness of the results to the
community and the degree of cooperation
expected (determined by earlier communi-
cations and a site visit by the PI).
(3) Levels of total dissolved solids and
competing ions, especially sulfate. High
total dissolved solids (>1000 ppm)
supplies are of particular interest for
comparing the desalting technologies
with the single contaminant removal
processes. Sulfate is of special interest
because it competes favorably with the
contaminants for adsorption sites on
alumina and anion exchange resins.
-------
(4) Presence of foulants such as iron,
manganese, silica, and organics. Attempts
will be made to choose water supplies so
that each of these foulants will eventually
be studied.
(5) Agreement by EPA (Drinking Water
Research Division) and state and local
governments that the supply to be studied
will produce useful and timely results.
General Specifications and
Layout
The research trailer is 3.2 m (10 ft) wide
rather than the usual 2.5 m (8 ft) to
provide a safe and comfortable working
environment for the field researchers.
Mobile offices and homes were available
in the 3.7-m (12-ft) width as the basic
unit, but they were rejected as being too
wide and flimsy for rugged use and
repeated moving. The basic trailer shell
was constructed of aluminum according
to UH specifications by General Truck
Body, Inc., of Houston, Texas, on a twin I-
beam, 8-wheel chassis. Though the 3.2-
m-wide (10-ft) trai ler is considered a wide
load, permits to move it are readily
obtained by contract haulers. Since 98
percent of the time the unit is in fixed-
locations and only 2 percent in transport,
the extra space, comfort, and safety
realized while it is stationary more than
compensate for the slight disadvantage of
transporting a wide load.
The pilot plant and the laboratory
sections have separate entrance doors,
and each entrance is provided with a
removable stairway and an attached
safety railing. In addition, the pilot plant
section has an extra wide (46-in., 1.17-m)
equipment door that doubles as an
emergency exit. A sliding door inside the
trailer allows the pilot plant to be isolated
from the analytical laboratory. Six flood
lights are mounted high upon the outside
of the trailer. These automatically turn on
at night to illuminate the entrances and
the area around the trailer to help
minimize vandalism. One double window
in the pilot plant and two single windows
in the laboratory provide ventilation,
natural lighting, and a view of the
surrounding area.
A special effort was made to fasten
rigidly all of the pilot plant and laboratory
equipment either to the floor or walls for
protection during transit. Unistrut chan-
nels were used throughout the trailer on
the walls and ceiling for mounting the
flow system piping and components and
the PVC electrical conduit and control
boxes. Also, weatherproof electrical
outlets, wall-switches, and control
enclosures were used throughout the
pilot plant and lab for protection against
possible process spills.
Corrosion-resistant plastics (PVC,
Plexiglas,* Teflon, polyethylene, and
nylon) were used wherever possible for
the pumps, valves, columns, and piping.
The widespread use of plastics would not
have been possible if organic rather than
inorganic contaminants were being
studied. Stainless steel, nylon, and fiber-
glass-reinforced plastics were used in the
reverse osmosis system, where pressures
in excess of 2760 kPa (400 psig) are
expected. Finally, the pilot plant compo-
nents were duplicated in the system
design wherever possible to provide
readily available replacements in the
event of failure.
The treatment processes are laid out in
a left-to-right flow scheme on one wall of
the pilot plant (Figures 3 and 4). The
feedwater, acid and base tanks, and
pumps are located on the left side of the
four processes, and the 200-L (55-gal)
polyethylene, treated-water and waste-
water collection tanks and pumps are on
the right. Larger 1160-L (300-gal) fiber
glass water and wastewater storage
tanks are mounted underneath on either
side of the trailer. Each of the treatment
processes is controlled from the graphic
control panel mounted on the process
wall above the reverse osmosis unit.
"Mention of trade names or commercial products
does not constitute endorsement or recommendation
for use.
Facing the control panel, the operator ^
stands within easy reach of both the \
process equipment and the graphic
controls. This arrangement is considered
an important safety and convenience
feature for the complex one-man research
operation.
Work and storage areas occupy the wall
across from the processes. The work
benches are used for equipment repair,
chemical preparation, and storage. Two
2.13-m-highx1.19-m-wide(7-ftx47-in.)
storage cabinets are provided for spare
parts, resins, alumina, and chemicals,
and the hand tools are kept in a rolling
tool chest. A safety shower and eyewash
fountain are located in the chemical
preparation area near the utility sink. The
analytical laboratory layout features
standard laboratory furniture with cast
epoxy bench tops and color-coded drawer
fronts (Figure 4). The laboratory also
serves as the project's field office. As
such, it contains a desk, file cabinet, book
shelf, and telephone.
The laboratory is equipped to analyze
raw and treated waters for pH, dissolved
solids, fluoride, chloride, nitrate, chrom-
ate, hardness, alkalinity, sulfate, and
silica by the usual wet chemical methods.
In addition, instrumentation has also
been provided to measure conductivity, ^
turbidity, silt density index (SDI), and Cl~, fl
F~, and N0a~ ions using electrodes.
Recently, an ion chromatograph was
added for routine analysis of the common
Figure 3. Pilot plant in mobile research facility. This view of the trailer is from rear to front, with
treatment processes on the left, work and storage areas on the right, and the door
opened to the laboratory in the center of the picture.
-------
pH Adjust Tank
(Hidden)
Ion Exchange
Columns
Activated
Alumna
.Agitator \\ Columns
Caustic\
Tank
Deep-Bed Filter
Control Panel
\Above RO Unit
iFeedpH
i Recorder
i Effluent
Electrodialysis Unit ,
Flowmeters
^Sampling Vessel
i Treated Water
Tank
Air-Conditioning
and
Heating Unit
Auto Sampler
Storage
Cabinet
'indow
Drying Rack Above Sink
Laboratory Benches
File Cabinet
.Exhaust Hood
Storage Shelves
Treated Water Pump
and
Waste water Pump
Work Benches
Rear Door
Feed Water Pump
Storage Compartment
Treated Water
and
Wastewater Tanks
(300 Gal. Each)
Pilot Plant Entrance
Power Panels
and Cabinet
Base Pump
Electrical Service
Figure 4. Interior layout of the UH/EPA drinking water treatment research facility showing both the pilot plant and laboratory equipment.
anions F , Cl , Br", N03 and SCV. A
rugged atomic absorption spectrophoto-
meter with a graphite furnace atomizer
will be provided whenever the research
trailer is being used to study arsenic or
selenium removal. Most of the water-
samples to be analyzed, especially
product water and regenerants, are
collected automatically by the automatic
sampler at times preselected by the field
researcher. Other grab samples of the
raw water and brine streams are collected
when necessary.
Pilot Plant Treatment Units
The primary components of the pilot
plant are the four treatment units.
Though they are interconnected, they
were designed to be operated one at a
time rather than simultaneously. Each
unit may be operated over a wide range of
feed and product water flow rates.
The 8-L/min (2-gpm) activated alumina
system is made up of two 20.3-cm-
diameter (8-in.) Plexiglas columns con-
taining 0.91 m (3 ft) of 28x48 mesh Alcoa
F-1 activated alumina. The columns may
be operated in series or parallel with
upflow or downflow exhaustion and
regeneration. The intended uses of the
alumina system are for fluoride, arsenic,
and selenium removal. Regeneration is
accomplished using dilute (0.25 N)
sodium hydroxide followed by acidification
with dilute (0.50 N) sulfuric acid. Spent
regenerants are either reused or neutra-
lized and disposed of locally.
The 8-L/min (2-gpm) ion-exchange
system is made up of two 25.4-cm-(10-
in.-) diameter Plexiglas columns typically
containing 0.91 -m (3 ft) of anion or cation
exchange resin. Single-bed or two-bed
ion-exchange processes may be simu-
lated with either upflow or downflow
exhaustion and regeneration. Single-bed
anion exchange with sodium chloride
regeneration is the intended method for
removal of nitrate, arsenic, selenium, and
chromate. As with the alumina system,
backwashing may be accomplished with
raw water or with treated water from the
storage tanks. In fact, with the exception
of the column diameters, the alumina and
ion-exchange systems are identical and
may be used interchangeably. Tanks and
pumps have been provided so that both
the ion exchange and alumina systems
may be regenerated or cleaned with acids,
bases, or salts in any conceivable se-
quence.
The reverse osmosis desalting system
is made up of two, different, standard,
hollow-fiber modules that may only be
operated one at a time. The DuPont
aramid module has a nominal product
water flow of 5.52 L/min (1.45 gpm) and
operates at a typical feed pressure of
2415 kPa (350 psig) with a typical product
water recovery of 50 percent. The Dow
cellulose triacetate module is considerably
larger, with a nominal product flow of
10.5 L/min (2.78 gpm), a feed pressure of
2415 kPa (350 psig), and 50 percent
recovery. Both units have typical overall
dissolved solids rejections in the range of
-------
95 to 98 percent and will be used to
remove any contaminant ion of interest.
As with the reverse osmosis and ion
exchange units, the reverse osmosis
product water may be stored temporarily
or disposed of locally with the rejected
brine. Proper pretreatment is known to be
the key to successful reverse osmosis
system operation, so the pilot plant has
been designed with means for dechlorin-
ation, polymer addition, deep-bed filtra-
tion, and cartridge filtration.
The reversible electrodialysis desalting
system was purchased as a complete unit
from the manufacturer, Ionics, Inc. The
nominal product-water flow rate is 1.31
L/min (0.35 gpm), and the recovery
varies from 50 to 80 percent, depending
on the amount of brine recycled. As with
the reverse osmosis unit, the reversible
electrodialysis unit will be used to study
the removal of all contaminant ions of
interest. Compared with reverse osmosis,
very little pretreatment is required for
reversible electrodialysis, and the system
operates with only cartridge prefiltration
and dechlorination using an activated
carbon filter. A membrane cleaning
system using acids and/or chelating
agents has been provided for cleaning the
reverse osmosis and reversible electrodi-
alysis membrances as required.
The full report was submitted in
fulfillment of Cooperative Agreement No.
CR806073 by the University of Houston
under the sponsorship of the U.S.
Environmental Protection Agency.
Dennis Clifford and Maheyar Bilimoria are with the University of Houston.
Houston, TX 77004.
Thomas J. Sorg is the EPA Project Officer (see below).
The complete report, entitled "A Mobile Drinking Water Treatment Research
Facility for Inorganic Contaminants Removal: Design, Construction, and
Operation," (Order No. PB84-145 507; Cost: $10.00. subject to change) will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
CM
U.S. GOVERNMENT PRINTING OFFICE: 1984-759-102/0891
------- |