Bakersfield Well GAC vs HiPOx paper 081605ECONOMIC AND OPERATIONAL COMPARISON OF GRANUALTED ACTIVATED
CARBON ADSORPTION AND ADVANCED OXIDATION TREATMENT
TECHNOLOGIES FOR THE REMOVAL OF MTBE AND TBA FROM DRINKING
WATER
By:
Nick Zaninovich, P.E.
Senior Engineer
South Tahoe Public Utility District
December2004
Abstract
The STPUD purchased, permitted and installed California’s first-ever advanced oxidation
process (AOP) treatment system utilizing O3/H2O2 to treat MTBE-contaminated ground
water for use as drinking water at its Arrowhead Well No. 3 (AH#3) in 2002. In July
2003, the STPUD shut down its Bakersfield and Country Club Wells, due to MTBE
contamination. This resulted in a loss of over 1,600gpm capacity for an already stressed
system. In addition, tert-butyl-alcohol (TBA) was discovered at a sentinel well within
2,500 feet of both AH#3 and Bakersfield wells. From June 2000 through July 2002, the
District retained Boyle Engineering of Fresno, California to conduct a study to compare
the efficacy of two types of granulated activated carbons (GAC). Although TBA
contamination can be present either independently or as a breakdown byproduct of
MTBE, the removal efficiency of TBA using GAC was beyond the scope of the pilot test.
In October 2003, STPUD engineers began conducting an extensive study to determine
the removal efficiency of GAC and AOP for low-level TBA and MTBE contamination.
This study incorporates STPUD’s first-hand experience with advanced oxidation
treatment of MTBE to compare the operational and economic aspects of both
technologies. Proposals were solicited from U.S. Filter of Oakland, California for their
proprietary coconut shell carbon treatment system, and from Applied Process
Technology of Pleasant Hill, California for their patented AOP system. The study utilized
full-scale design, construction, and operating cost data from the existing AH#3 AOP
facility, while the GAC option was estimated based upon known and estimated hard
costs, information provided by U.S. Filter, Boyle Engineering, and other water utilities.
Extensive literature research was also conducted to help determine infrastructure
requirements and operational peculiarities of each technology. Cost data were based
upon 15-year net present value for capital, operational, and maintenance expenses
based upon nine (9) probable occurrence scenarios of MTBE and TBA. Due to the
environmental sensitivity of the Lake Tahoe basin, site impacts were also a significant
consideration. This report provides the quantitative and qualitative results of the
comparison.
Introduction
The South Tahoe Public Utility District (District) is located on the south end of Lake
Tahoe, California, and provides drinking water to approximately 13,000 residential and
commercial customers from ground water wells. The base population within the
District’s service area is approximately 28,000, with tourists creating occasional
population swells to over 150,000 during the peak seasons.
1
In 1998, methyl-tertiary-butyl-ether (MTBE) was discovered in several of the District’s
wells. Consequently, 14 of the District’s 35 wells were shut down either for direct or
potential contamination. The District’s Board of Directors adopted a “zero tolerance”
policy for MTBE contamination, establishing a 0.2ppb detection limit standard for
requiring shutdown of wells.
In 1999, the District filed suit against local service station owners and petroleum
refineries to cover costs related to present and future remediation expenses due to
leaking fuel tanks and piping systems at service stations. In August 2002, defendants
settled with the District for a total of $69 million.
During the lawsuit, the District conducted pilot testing to determine the effectiveness of
granulated activated carbon (GAC) for the removal of MTBE. The testing was performed
using Calgon Carbon F-600, and U.S. Filter CC-602 carbons at the District’s Clement
Well site, using water from the Tata Lane Well. The pilot study determined that GAC is
effective at adsorbing low-level MTBE contamination, with the U.S. Filter GAC being
more effective than the Calgon Carbon GAC. Additionally, a proprietary AOP process
from Applied Process Technology (APT) of Pleasant Hill, California was also pilot tested
between April and August 2000.
In 2001, the District determined it was necessary to treat the Arrowhead Well No. 3
facility to meet customer demand. Due to anticipated higher-level contamination, and
Board member concerns over potential GAC adsorber breakthrough, APT’s AOP
process was further evaluated.
Due to ideal water chemistry and portability, the AOP was the apparent best solution,
guaranteeing MTBE destruction and low byproduct production. The AOP facility was
permitted and constructed in 2002, and has proven successful at MTBE destruction.
This is the first application of AOP technology for the treatment of MTBE in drinking
water in the State of California.
Discovery of MTBE at the Bakersfield Well
In April 2003, MTBE contamination was discovered at the District’s Bakersfield Well
(located approximately 4,000 feet north and down-gradient of the Arrowhead Well No. 3
facility). Additionally, MTBE contamination was discovered at the District’s Country Club
Well, located approximately 7,000 feet north and also downstream of the Arrowhead
facility.
Continued monitoring of the Bakersfield Well during the summer of 2003 indicated
MTBE contamination levels varying from 0.2ppb to 0.6ppb. Monitoring of the District’s
Sioux Street sentinel well, located approximately equidistant between the Bakersfield
and Arrowhead wells, showed levels of MTBE varying from 0.6ppb to 5.8ppb from May
through September 2003. The Sioux well also showed levels of TBA on two different
occasions, varying from 5.7ppb to 6.3ppb. Xylene and toluene were also present in
several samples. Table 1.1 summarizes the average chemical and physical
characteristics of the well’s water for 2003.
Since TBA contamination can be present either independently or as a by-product of the
natural degradation of MTBE, the District is uncertain as to the exact source of its
2
Table 1.1
Average 2003 Physical and Chemical Characteristics of
Bakersfield Avenue Well Water
Temperature 11.2° C
PH 8.9
Turbidity 0.04
TDS 101
Alkalinity 60.2
Hardness 28.2mg/l as CaCO3
Langelier Index 0.05
TOC 0.24mg/l
Cl−9.53mg/l
NO2 <0.010mg/l
NO3 0.42mg/l
Fl 0.17mg/l
SO4 2.30mg/l
Ca 10.43mg/l
Cu 20mg/l
Fe 50mg/l
K 2.00mg/l
Mg 0.67mg/l
Mn 20mg/l
Na 21.4mg/l
Br 0.010mg/l
P (total) 0.036mg/l
Zn <10mg/l
As 11.48mg/l
Ag <10µg/l
Ba <100µg/l
Pb <2.50µg/l
Uncorrected Gross α 12.82pCi/l
U 6.07pCi/l
Corrected Gross α 6.75pCi/l
Ra-226 0.00pCi/l
Rn 234pCi/l
3
Presence. Soil samples from a nearby gasoline station suspected of contaminating the
aquifer did not reveal TBA. However, due to TBA’s relationship to MTBE and known
carcinogenicity, the District held the position that neither TBA nor MTBE were
acceptable for drinking water at any detectable level.
As a result of the contamination, both the Bakersfield and Country Club wells were shut
down in September 2003. These shutdowns resulted in an additional loss of 1,500gpm
and 200gpm, respectively, from the system’s production capability.
Decision to Treat and Basis of Study
Due to the substantial loss of production capability, the District decided to provide
treatment for the Bakersfield Well, while opting to leave the Country Club Well shut down
due to its much lower productivity and resultant high cost-to-benefit ratio.
Since MTBE levels were expected to be much lower than what were anticipated at the
Arrowhead No. 3 Well, GAC appeared to be the most viable treatment option. However,
since none of the GAC pilot testing considered TBA removal, District engineering staff
performed a comprehensive study to determine its efficiency at TBA removal, and
compared the effectiveness of GAC adsorption to AOP destruction of both MTBE and
TBA.
The study consisted of nine different contamination scenarios, each scenario
representing assumed possible combinations of MTBE and TBA based upon correlating
contaminant levels at the Sioux sentinel well, which is under the influence of Bakersfield
Well (see Tables 1.2 and 1.3). These scenarios were also based on qualitative analysis
of historical data of area wells.
Description of HiPOxTM System
APT’s proprietary system consists of a stainless steel pipe plug flow reactor, with one
sidestream injection port for H2O2 on the influent end followed by numerous injection
ports for ozone. A final injection port on the reactor is for NaOCl to destroy any residual
H2O2 and to provide free chlorine residual in the distribution system. A static mixer
immediately follows each injection point. The reactor and ancillary equipment are
contained in one of two steel sea cargo containers. The second container houses the
process control equipment and chemicals. System hydraulic controls are housed in an
adjacent structure. The H2O2 and O3 react together in water to form the hydroxyl radical
(OH-) through the following reactions:
H2O2 + H2O → HO2− + H3O2+
HO2- + O3 → • HO2 + O3-
2O3- + H2O2 → 3O2 + 2 • OH
Through a series of progressive, intermediate reactions within the reactor, volatile
organic compounds (VOC’s) are broken down into carbon dioxide and water. The
carbon dioxide is off-gassed at the top of the reactor through an air release valve,
represented by:
4
MTBE O2 → Intermediates O2 → CO2 + H2O
• OH • OH
Table 1.2
2003 Contaminant Levels at the Sioux Sentinel Well
DATE MTBE TBA
May 14 0.56µg/l 0
June 5 2.4µg/l 6.3µg/l
June 23 2.6µg/l 5.7µg/l
July 17 3.9µg/l 0
August 7 3.2µg/l 0
September 22 2.9µg/l 0
October 28 2.7µg/l 0
Table 1.3
Nine Scenarios of Possible Combinations
of MTBE and TBA Contamination
Scenario MTBE, µg/l TBA, µg/l
1 0.5 0
2 2.0 0
3 5.0 0
4 0.5 2.0
5 2.0 2.0
6 5.0 2.0
7 0.5 6.0
8 2.0 6.0
9 5.0 6.0
Due to the limited pressure capabilities of the ozone generators to inject ozone into the
reactor, a pressure reducing/pressure sustaining valve is used at the wellhead to control
influent pressure and provide for steady pressure conditions. During system startup,
treated effluent is initially flushed to waste through a solenoid-operated control valve until
two sets of analyzers show that the water contains no residual ozone or H2O2, and that
there is a residual of 0.5 to 1.5 ppm of free chlorine. Once the conditions are met, the
control valve is called to close, while the booster pump is called on and the PLC/PID-
controlled flow control valve is opened.
While the Arrowhead installation is designed for a constant rate of flow at 800 gpm, the
Bakersfield installation is designed for variable flow, controlled via the treatment unit
PLC and a cascade control to the PID controller on the final flow control valve in
conjunction with a VFD on the well pump motor. Flowrates can be 1,000, 1,250, or
1,500 gpm depending upon the distribution system storage tank level.
5
With a VFD driving the pump motor at Bakersfield Well, it was unnecessary to change
out the existing well pump and motor, in contrast to the Arrowhead system required
installation of a lower-head producing pump and motor in the well. The upstream side
of the pressure reducing/pressure sustaining valve sustains 50 psi inlet pressure
regardless of flowrate and quiets surges during increases in the flow. The downstream
pressure is set to sustain a constant 40 psi for each flow setting.
After the water leaves the reactor, it is boosted up to distribution system pressure
through a centrifugal pump, with the final control valve controlling flowrate into the
system. If flowrate adjustment is required, the adjustments are made via the PLC in a
timed fashion to allow the VFD and chemical feeds time to properly adjust while
maintaining the desired hydraulics and water quality parameters throughout the change.
Construction of Arrowhead Well No. 3 Advanced Oxidation Treatment Facility
During the course of the District’s lawsuit, all available technologies for treating or
removing MTBE from water were evaluated. The only two technologies that were found
to reliably treat for MTBE contamination were AOP and GAC.
Due to the District’s Board of Directors desire to guarantee reliable, non-detectable
amounts of MTBE in the drinking water system, destructive technology was the
preference. Since the proprietary AOP “HiPOxTM” (hyperoxidation) process was the
most expensive option, it also assisted in maximizing the claim against the polluters.
This allows the District to pursue potentially expensive, yet unforeseen technologies for
future MTBE mitigation.
Additionally, the HiPOxTM system appeared to be “portable” in comparison to the GAC
option, being contained in two steel sea cargo containers that can be easily moved by
truck and crane. While the HiPOxTM system is designed to be portable, it was not
designed to be self-contained with all amenities required to treat drinking water, nor was
it designed for Lake Tahoe’s cold, snowy winters.
The District purchased the HiPOxTM unit in used condition from APT. The unit was
originally installed in Burbank, California (Southern California), where freezing
temperatures are seldom, if ever, seen. Such climate allows for piping to be installed
above ground on temporary installations. Under this scenario, one of the features of the
unit was that it could be set up on a site within five days’ time, if the need arose. Similar
units are typically used to treat various types of organic ground water contaminants for
relatively short durations, discharging treated effluent to either a storm drain or sanitary
sewer system. The portability was an important feature for the District, since there was
speculation that the unit could be moved to other sites once the contaminated ground
water plume cleared at the Arrowhead site.
In order to adapt the unit to Tahoe’s freezing climate, both the reactor and control units
had to be retrofitted with insulation and provided with wall-mounted heating and air
conditioning units. Piping that can normally be left exposed in warmer climates had to
be buried underground for this installation, which proved to be an extensive undertaking.
The control unit is housed in an 8’x20’ container. This unit contains the programmable
logic controller (PLC), a circuit breaker panel, hydrogen peroxide storage tank and
metering pumps, two ozone generators, and an emergency shower/eye wash unit with a
small hot water heater. The process unit houses the reactor (consisting of
6
approximately 400 linear feet of serpentined stainless steel 10” piping with 18 injection
points/static mixers), and meters for treated effluent flowrate, residual ozone, hydrogen
peroxide, and sodium hypochlorite. The units are placed side-by-side with a conduit
raceway constructed between them.
The existing well control building that originally housed all of the electrical and flow
controls prior to the discovery of MTBE contamination was converted for use as a
booster pump building. Additionally, a small valve building was constructed due to the
space limitations inside the control building.
The flowrate of the treatment unit is designed for a steady 800 gpm, although the reactor
itself has a capacity of up to 1,200 gpm. A requirement of the installation was to be able
to flush substantial waste volume from the process startup to the sewer system.
Unfortunately, the sewer lines near the Arrowhead site occasionally surcharge. The
sewer was estimated to be capable of accepting a flowrate from the site of 150 gpm,
requiring an attenuation of approximately 650 gpm. Therefore, a 10,000 gallon waste
detention tank was placed on the site to act as a peak reducer.
After six seventy-hour weeks of underground piping, grading, concrete, and paving work,
it became readily apparent that the HiPOxTM units were not nearly as portable in the
Tahoe Basin as they were in Southern California. Considering the more short-term
installations and the warmer, drier climate of Southern California, the units can be set
with minimal anchorage, and all piping can be placed above ground. Further, no
separate waste flow lines or booster pumping was required for such installations, since
all flows went to waste.
Unfortunately, the District was unable to obtain any bidders for the project due to the
concerns related to time constraints and risk due to the new technology. However,
several contractors were willing to work on a time and materials basis under
superintendence of one of the District’s engineers to complete underground and site
work, leaving the mechanical work for District in-house forces to complete.
HiPOxTM treatment system design, purchase of the unit, and construction expenses for
the Arrowhead Well No. 3 Treatment Project are summarized in Table 2.1, with
operating expenses summarized in Table 2.2 (both in 2002 U.S. dollars).
Bakersfield Well Treatment Technology Selection
With the District’s experience with the HiPOxTM system at Arrowhead Well, there were
staff concerns, chiefly among the system operators, that the HiPOxTM system was too
complex. This position led to a reconsideration of the viability of GAC treatment.
However, with the concern about the potential presence of TBA, but the lack of pilot
study information on GAC effectiveness for TBA removal, the District was unsure of the
GAC option. As a result, the District decided to evaluate the economics, operation, and
maintenance perspectives of both GAC and HiPOxTM technologies.
Capital, operation & maintenance costs were estimated and evaluated for a 15 year net
present value (NPV). With a lack of first-hand knowledge of full-scale GAC operations,
staff investigated GAC installations and related operational issues in the cities of Fresno
and Clovis, California, where such units are being used extensively to treat for
dibromochloropropane (DBCP) contamination. Staff also performed literature research
7
Table 2.1
Arrowhead Well No. 3 Treatment
Project Expenses
(2002 Dollars)
Table 2.2
Average Monthly Chemical Consumption,
Production, and Operating Costs
(2002 Dollars)
Mo. Monthly
Production
(Mgal)
1O2Usage
(ccf)
O2 unit
Price
($)
H2O2Usage
(gal.)
H2O2Unit
Price
($)
NaOCl
Usage
(gal)
NaOCl
Unit
Price
($)
Electrical
Usage
(KWH)
Electrical
Cost
($)
Tele-
Phone
($)
Engineering
Labor
($)
Pump
Crew
Labor
($)
3Total
Cost/
Mo.
($)
3Total
Cost/
Gal.
($)
Oct.
‘02
34.8 57,700 0.0055 111 5.00 603 1.23 32,000 3,180 15 510 2,790 8,619 0.000248
Nov.
‘02
26.6 55,800 0.0055 88 5.00 519 1.23 31,000 3,080 15 490 2,700 8,180 0.000308
2Dec.
‘02
12.9
(28.6)4 26,000 0.0055 40 5.00 257 1.23 14,000 1,390 15 230 1,260 4,050
(8,968)4 0.000315
1O2 tank bleeds off frequently due to low draw. Exact system consumption is therefore unknown.
Design and consulting $107,822
HiPOx system purchase $696,018
Construction contracts (time and materials basis) $268,887
Materials and supplies (District purchases) $180,653
District labor:
Engineering $70,099
Pump Crew $30,193
Heavy Maintenance Crew $19,077
Electrical Crew $2,204
Underground Repair Crew $1,722
Laboratory $585
Permits $3,255
Legal $528
Lab Monitoring $20,999
Other/Misc. $9,468
TOTAL $1,411,510
5 Monthly average for 3 months of operation: $8,589 0.000290
214 days of operation only. 3Includes $510/mo. O2 tank lease and maintenance fee. 4Extrapolation to 31 days of operation 5Includes adjustment of Dec. ’02 data to 31 days of operation
on both technologies related to contaminant removal and operational efficiencies.
Proposals were solicited from both U.S. Filter and Applied Process Technology for the
competing systems. Engineering staff performed planning-level designs for supporting
infrastructure and operation and maintenance cost estimates. While the District had
8
precise costs for the operation of the Arrowhead HiPOxTM system, the operational
process and site requirements of each technology were substantially different, and had
to be determined in detail. The supporting infrastructure requirements for GAC were
completely different than for HiPOxTM. Operational and maintenance costs were
estimated by information obtained from Fresno and Clovis, and U.S. Filter. Ease of
operation and limited maintenance requirements were evaluated qualitatively.
While the HiPOxTM system seemed complex to the District’s operators, GAC was
determined to have some critical, but less obvious operational concerns as well. One
concern was regarding competitive adsorption of MTBE and TBA with each other, and
with other constituents in the water, particularly arsenic and radionucleides. The
concern with arsenic was not only related to adsorption, but also desorption, and
potential hazardous waste material liability (ie “cradle-to-grave”) issues. While there
appeared to be no current mandates related to this concern, fear of potentially changing
or new future regulations were considered. Radionucleide accumulation was a concern
of the California Department of Health Services. The Bakersfield Well water contains
uranium, which can adsorb onto the carbon, then deteriorate to lead IV, which could
accumulate inside the vessels and potentially desorb en mass.
Second, the operational scheme of the well has historically been a tank level call on/off,
with scaling to allow for variable frequency drive (VFD) pump motor control. With GAC,
the flowrate variability could potentially alter the mass transfer zone (MTZ), resulting in
finished water quality inconsistencies and difficulties for system operations. Due to the
long MTZ required to adsorb MTBE, low to moderate loading rate, and long empty bed
contact time (EBCT) required, two sets of three 30,000 lb. adsorbers operating in parallel
were required.
Finally, biogrowth inside the adsorbers was a concern due to the potential for substantial
headloss through the vessels, and again the potential for mass desorption. Although no
solid evidence exists, it is theorized that biogrowth occurred during the GAC pilot testing,
which possibly could have extended breakthrough time by biologically degrading the
MTBE.
While the capital cost for the HiPOxTM system was decidedly more than that for GAC, the
15-year NPV (assuming an annual inflation rate of 3%) for ongoing operations was less
with consideration for TBA. As TBA levels increase, carbon consumption goes up
considerably. Not only does this add expense to the treatment for carbon changeouts,
but can result in considerable out-of-service time and operator labor for changeouts and
flushing. Flushing is a concern due to limited sewer line capacity within the area of the
project, which would require a large detention facility to prevent overload of the sewer
system, increasing operator time and system out-of-service time considerably. In Lake
Tahoe’s sensitive environment, land coverage is a primary concern. The land coverage
required for six (6) 30,000 lb. GAC vessels is considerably more than what was required
for a HiPOxTM system. With a detention facility installed with the system, site coverage
would have been excessive in contrast to local regulations. Additionally, with the well
located in a quiet residential neighborhood, the GAC plant would have had a large
profile, compared to the discrete profile of the HiPOxTM system.
With the potential for an unknown breakthrough of MTBE or TBA on one of the District’s
largest wells, the Board’s position was viewed as more favorable for the destructive
9
technology of the HiPOxTM process. Table 3.1 presents a side-by-side comparison of
pros and cons for each of the two competing technologies.
Table 3.1
BAKERSFIELD WELL HiPOx AND GAC TREATMENT ALTERNATIVE COMPARISON TABLE
CATEGORY HiPOx GAC
A. Required Infrastructure
Site Coverage Minimal (16'x20') Substantial (50'x50' +/-)
Profile Low (12' +/-) Tall (25'-30')
Enclosure Self-contained Requires constr. of large enclosure
Electrical Substantial additional load Negligible additional load
Mechanical More control valves, less piping More hand valves and piping
B. Reliability
Capacity 1,800gpm maximum 1,500gpm maximum
Mechanical Subject to occasional failure Less prone to failure
(automatic shutdown) (no automatic shutdown)
More frequent, but short periods of Less frequent, but longer periods
down time of down time
No negative water quality impact Potential for biogrowth during
during preriods of shutdown periods of shutdown
Process Guaranteed destruction of MtBE/TBA Subject to breakthroughs and
sudden release of accumulated
compounds
Chemical & Biological Could create various byproducts May become bioreactive
(other oxygenates, bromates, and
unknowns)
No hazardous waste generated Subject to accumulation of other
compounds, especially radio-
nucleides and degradation
daughters
Requires routine testing for Potential for hazardous waste
Oxygenates disposal issues
Arsenic synergy (change of valence) Potential for taste & odor issues
10
Table 3.1, Continued…
C. Operations & Maintenance
Staff time Substantial (44hrs. +/- per month) Minimal (15hrs. +/- per month)
Requires involvement from both Requires only Pumps crew
Pumps and Electrical crews
Controls Complex - electronic Basic - manual
Materials handling Oxidizing chemicals (H2O2, O3) Minimal, generally non-hazardous
On-site generation of oxygen and
ozone (as needed - no bulk storage)
Waste flows Approximately 3,000gal. Approximately 200,000gal.
at each startup at each changeout, approximately
11,000gal. per vessel per backwash
(66,000gal. for six vessels)
Requires on-site waste tank, minimal Requires portable or permanent
operator time settling tank, substantial operator
time at each backwash/changeout
While the HiPOxTM process is more technologically encumbering, it is capable of reliably
destroying both MTBE and TBA with only modest increases in chemical dosage for
modest increases in contaminant levels. However, bromide can be an issue with the
AOP process due to the potential for bromate formation. Fortunately, the well’s bromide
levels are well below the ability to form bromate in excess of the MCL.
Other primary concerns related to the HiPOxTM system is the formation of undesirable
intermediate byproducts. However, the District’s experience from laboratory testing with
the Arrowhead HiPOxTM system effluent has shown that this concern may not be as
merited as first thought for low-level contamination. However, not every potential
byproduct is, or can be monitored.
Capital and O&M Estimation – GAC
Table 3.2 summarizes the approximate capital expenses associated with the proposed
GAC system. The system required six 30,000 lb. vessels, connected in two series’ of
three vessels each. U.S. Filter included the cost of the vessels, connective piping, and
the first load of carbon in their proposal. Prefilters were an option added by the District
for the purpose of extending adsorber life and reducing headloss. The District would
have been obligated to construct a concrete slab to support the units, and would have
had to provide a heated enclosure due to freezing temperatures.
11
Table 3.2
GAC OPTION - ENGINEER'S ESTIMATE
Direct Capital Cost (2003 Dollars)
DETAIL QUANTITY UNITS
UNIT
PRICE EXTENSION
TREATMENT SYSTEM
US Filter 2x3x30,000lb
GAC 1 EA 601000 601,000
Pre-filters 2 EA 18000 36,000
SITE WORK
Mobilization/demobilization 1 LS 35000 35,000
Grading/tree removal 1 LS 10000 10,000
Concrete Foundation 1 EA 52000 52,000
Access pavement 1200 SF 7.5 9,000
Revegetation 1 LS 5000 5,000
Backwash tank/pad 1 EA 25000 25,000
Backwash line, valves,
fittings 180 LF 80 14,400
Permitting / Coverage 1 LS 10000 10,000
Landscaping 1 LS 5000 5,000
MECHANICAL
Piping and valves 1 LS 25000 25,000
STRUCTURAL
Building 2500sf 1 LS 260000 260,000
Structural design consultant 1 LS 5000 5,000
Subtotal 1,092,400
Contingency 10% 109,240
TOTAL: $1,201,640
Estimate includes contractor markup
The vessels would have required a pad measuring approximately 40’ x 60’, with an
enclosure tall enough to accommodate the vessels (approximately 24’). Additionally, a
paved access driveway would have been constructed to allow truck access for carbon
changeouts. U.S. Filter would have provided the delivery and setup under their
proposal. Modifications of the existing wellhead control piping would have been
undertaken by the District, but would not have been extensive. A backwash tank and
pad would have been required due to the high backwash volume and flowrate in contrast
12
13
to the relatively low flow capacity of the adjacent sewer system. The District evaluated
the use of a portable trailer mounted tank to conserve site coverage. Components are
included in the estimate for backwash piping to the waste tank. Due to the site
requirements of the system exceeding the allowable limits set forth by the TRPA,
additional coverage would have been required to be purchased.
Table 3.3A gives approximate O&M costs based upon information provided by U.S.
Filter, City of Fresno, and City of Clovis. To develop approximate staff time expenses,
the District’s operators were consulted in detail to determine estimated time to perform
required tasks for each of the nine scenarios.
Lab testing involves primarily monitoring for breakthrough, and requires increasing
testing frequency as the MTZ moves through the carbon bed toward breakthrough. An
average of $249 per month was included for approximate power consumption required
for building heat.
Table 3.3B provides the NPV for the carbon consumption costs for each of the nine
scenarios under the react and return option, which appeared to be more economical
than virgin carbon at each changeout. Under react and return, the carbon can be
regenerated up to two times before it is no longer economical to do so. With each
regeneration, efficiency of the carbon adsorptive capability is reduced significantly which
reduces the duration of the changeout cycle.
Capital and O&M Estimation – HiPOxTM
Table 4.1 summarizes the approximate capital expenses associated with the HiPOxTM
system. The system required two custom modified cargo containers (8’x28’ each), with
concrete spread footings. As opposed to the Arrowhead system, where a bulk liquid
storage tank was used to supply the ozone generators with oxygen, a pressure swing
adsorption (PSA) generation system was used for Bakersfield Well. The desirability of
the PSA unit over the oxygen storage tank was due to the near proximity of a schoolyard
to the site, and reducing required site coverage since the PSA unit could be housed
within the control facility. While the PSA option added substantial cost to the project,
safety and security concerns could be adequately addressed.
Due to the size of the Bakersfield Well control building, most piping was set above
grade, with a short pipe chase constructed to allow for interconnection to the treatment
process containers, which are set directly behind the existing control building. The
mechanical controls and piping were redesigned inside of the existing building,
drastically reducing the amount of underground piping required, in contrast to the
Arrowhead Well, which is underground-intensive. A component was included in the
estimate for T-111 siding, trusses, faux windows, and a composition roof to be
constructed onto the sides and top of the cargo containers to enhance aesthetics.
Again, due to coverage concerns and potential for future arsenic treatment requirements
at the site, the flush tank was designed out of 42” Class IV RCP pipe, which is buried
underground, and can accommodate having the arsenic treatment unit installed on a
concrete pad constructed over the top of it.
While a new well pump was originally considered, it was determined to be unnecessary
due to the ability to apply a pressure control scheme to the VFD for the existing pump.
14 Table 3.3AGAC OPTION - ENGINEER'S ESTIMATE O&M Costs District Operational Staff Time: Lab BuildingMonthly ScenarioDays B/TRoutineChangeoutProblemsFiltersTestingPowerMaint.TotalNPV $/mo. $/mo. $/mo. $/mo. $/mo. $/mo.$/mo. $/mo. (3%,15yr) 1 1980750922001088592491252,384$345,1792 10757501703681081,0312491252,802$405,6743 7197502545511081,2372491253,274$474,1104 14010001,3042,8291081,4842491257,099$1,027,9835 12810001,4263,0941081,5592491257,561$1,094,8266 11610001,5733,4141081,6362491258,106$1,173,8137 9612001,9014,1251081,7182491259,427$1,365,0258 9012002,0284,4001081,7182491259,828$1,423,1989 8412002,1734,7141081,71824912510,287$1,489,683
Table 3.3B
GAC OPTION - ENGINEER'S ESTIMATE
Carbon Costs (React-Return Option)
1500gpm 500gpm 750gpm NPV1
$/yr $/yr $/yr (3%,15yr)
8,170 3,468 5,819 $69,467
15,063 6,384 10,724 $128,016
22,595 9,545 16,070 $191,843
140,649 59,680 100,165 $1,195,757
162,297 64,121 113,209 $1,351,482
185,471 68,664 127,068 $1,516,924
196,356 85,484 140,920 $1,682,294
221,329 89,630 155,480 $1,856,104
239,553 94,230 166,892 $1,992,340
1NPV based upon average production rate of 750gpm
However, the booster pump was still required to get the treated effluent up to distribution
system pressure. A rubber-bladder pneumatic surge tank was installed on the suction
side of the booster pump to protect it from flow transients during startup and shutdown,
and while control valves are being opened or closed. Due to the electrically and control-
intensive nature of the project, a control and instrumentation engineer was used for that
portion of the design.
Table 4.2 provides the estimated O&M NPV for a 15-year operation. In contrast to the
GAC, where costs for O&M are highly dependent upon the contaminant scenario,
HiPOxTM is less so. This is due to the ability to make simple modifications to the
chemical feed rates through the PLC, based upon the known required oxidant levels to
treat specific levels and types of contamination. Components were also included for
estimated repair and routine maintenance for the chemical feed pumps.
Table 4.3A provides the approximate 15-year NPV for electrical consumption of the PSA
unit ozone generator, booster pump, and the chemical cost for hydrogen peroxide. Cost
to run the well pump itself was not included in either the GAC or HiPOxTM estimate, since
the cost would be approximately the same for each option. Table 4.3B provides the
approximate electrical consumption of the PSA unit on a monthly basis. This expense
can be contrasted to the liquid storage tank expenses, presented in Table 2.2.
Conclusions
Table 5 shows the total 15-year NPV for each of the two options considered, for each of
the nine different contamination scenarios. The NPV’s shown include all estimated
capital and O&M expenses previously presented herein. Figure 1 illustrates the
differences between the GAC and the HiPOx options using the data provided in Table 5.
15
Figure 1 is a graphical interpretation of Table 5, showing GAC to be more economical for
low level MTBE contamination but less effective for TBA. HiPOx appears to be
substantially more economical for the treatment of MTBE and TBA.
Table 4.1
HiPOx OPTION ENGINEER’S ESTIMATE
Direct capital cost (2003 dollars)
DETAIL QUANTITY UNITS UNIT PRICE EXTENSION
($) ($)
TREATMENT SYSTEM
APT HiPOx Unit 1 EA 743000 743000
PSA O2 system 1 EA 90000 90000
SITE WORK
Mobilization/demobilization 1 LS 35000 35000
Grading/tree removal 1 LS 5000 5000
Underground water lines 1 LS 11000 11000
Concrete footings 2 EA 5000 10000
Access pavement 1500 SF 7.5 11250
Set units in place 1 LS 5000 5000
Waste line & facilities 1 LS 38000 38000
Revegetation 1 LS 5000 5000
Waste tank/pad 1 EA 25000 25000
MECHANICAL
New well pump 1 EA 8500 8500
Piping and valves 1 LS 57000 57000
Booster pump/pad 1 EA 15000 15000
Water quality analyzers 1 LS 7500 7500
Meters & gauges 1 LS 17000 17000
NaOCl tank 1 EA 3500 3500
Surge tank 1 EA 13,000 13000
Consultant electrical
design 1 LS 31,750 31750
STRUCTURAL
Façade work 1 LS 8000 8000
Pitch roof 1 LS 15000 15000
Building modifications 1 LS 2000 2000
SUBTOTAL: 1140258
10% CONTINGENCY: 114026
Estimate includes contractor markup TOTAL: $1,254,283
16
17
Table 4.2
HiPOx OPTION ENGINEER'S ESTIMATE - O&M NET
PRESENT VALUE FOR 15 YEAR OPERATION
DETAIL BASIS OF NPV
COST (3%, 15yr.)
OPERATIONS
Staff time $3211/mo. (2003) $464,970.37
Water quality testing
$1,123/mo.
(2003) $162,616.54
Sodium Hypochlorite $679/mo. (2003) $98,322.92
MAINTENANCE
Ozone generator $4,000/5yrs. $8,994.26
Dosing pumps $1,000/2yrs. $5,564.57
TOTAL: $740,468.66
NOTE: O&M costs are assumed nearly constant regardless of
contaminant level present for scenarios evaluated
The District generally concluded from this study that HiPOx offered the best efficiency for
the contaminants of concern while effectively meeting environmental concerns. While
GAC was not determined to be desirable for this particular application, the District
considers it viable treatment for MTBE where TBA has not shown a presence.
While there were very specific environmental considerations for the Tahoe installations
that impacted the economics of both technologies, it should be noted that similar
installations of either technology in less comparable geographies may yield different
economics than what is represented in this study. Since the District may relocate the
HiPOx facilities once contamination clears at either of the two sites, no salvage value
was included in theses estimates.
References and Resources
Personal Interviews:
Mark Reitz, P.E., Managing Engineer, Boyle Engineering Corporation, Fresno, California
Brock D. Buche, Water Division Engineer, City of Fresno, California
Ken Heard, Operations Supervisor, City of Fresno, California
Lisa Kohen, Assistant Public Works Director, City of Clovis, California
Cathy Lee, State of California Department of Health Services, Sacramento, California
Ivo Bergsohn, Hydrogeologist, South Tahoe Public Utility District, South Lake Tahoe, California
Phillip Torney, Pump Operations Supervisor, South Tahoe Public Utility District, South Lake Tahoe, California
Glenn Roderick, Water System Operator, South Tahoe Public Utility District, South Lake Tahoe, California
Table 4.3AHiPOx OPTION ENGINEER’S ESTIMATE – CHEMICAL AND ELECTRICAL CONSUMPTION Net Present Value for 15 - Year Operation SCENARIO MTBE TBA 1H2O2 1,2OZONE GEN. BOOSTER PUMP H2O2 ELECTRICITY TOTAL NPV (ppb) (ppb)(gal/yr)(KWH/yr) (KWH/yr) ($/yr)($/yr) ($/yr) (3%/yr)1 0.2 012482149012000062401414920389 243,4032 2022883939012000011440 1593927379 326,8493 5028084843312000014040 1684330883 368,6834 0.52291550219120000145751702231597 377,2025 2238496631712000019245 1863237877 452,1706 5244767703812000022380 1970442084 502,3947 0.56531091413120000265502114147691 569,3368 26635010936512000031750 2293754687 652,8449 56700412033412000035020 2403359053 704,9761H2O2 usage prorated at 66.7% of maximum consumption for annual flow variation 2Electricity provided for ozone generator and PSA system, prorated at 66.7% of maximum consumption for annual flow variation from values shown in Table 4.3B 18 References and Resources (cont’d.) Proposals: Applied Process Technology, Pleasant Hill, California (Prepared by Steve McAdams, Vice President of Manufacturing) U.S. Filter - Westates, Oakland, California (Prepared by Grant King, Field Sales Engineer) Technical References: Water Quality and Treatment, Fifth Ed., Chapter 13 – Adsorption of Organic Compounds, American Water Works Association, Letterman, et al, MCGraw-Hill, New York, 1999 Water Treatment Plant Design, Third Ed., Chapter 14 – Activated Carbon Processes, American Water Works Association & American Society of Civil Engineers, McGraw-Hill, New York, 1998
Table 4.3B
HiPOx OPTION ENGINEER’S ESTIMATE – P.S.A. SYSTEM ELECTRICAL CONSUMPTION
(2003 Dollars)
O3
% OF
120lb/d PSA O3 TOTAL TOTAL
SCENARIO (lb/d)
Max. O3 output (KWH/mo.)(KWH/mo.)(KWH/mo.) (KWH/yr.)
1 21.6 0.18 2259 426 2685 32219
2 39.6 0.33 4141 780 4921 59056
3 48.7 0.41 5093 958 6051 72613
4 50.5 0.42 5281 993 6274 75291
5 66.7 0.56 6975 1310 8285 99426
6 77.5 0.65 8105 1520 9625 115499
7 91.9 0.77 9611 1810 11421 137051
8 110 0.92 11504 2160 13664 163966
9 121 1.01 12654 2380 15034 180410
Table 5
Engineer's Estimate - GAC and HiPOx Treatment Options for Bakersfield Well
Total 15 year Net Present Value - 3% net discount rate
MTBE1 TBA2 GAC NPV HiPOx NPV GAC HiPOx
SCENARIO (ppb) (ppb)x(i) x(i) p(i)3 x(i)p(i) x(i)p(i)
1 0.5 0.0 $1,616,286 $2,238,155 0.250 $404,072 $559,539
2 2.0 0.0 $1,735,330 $2,321,601 0.180 $312,359 $417,888
3 5.0 0.0 $1,867,593 $2,363,435 0.160 $298,815 $378,150
4 0.5 2.0 $3,425,381 $2,371,954 0.150 $513,807 $355,793
5 2.0 2.0 $3,647,948 $2,446,922 0.125 $455,994 $305,865
6 5.0 2.0 $3,892,377 $2,497,146 0.075 $291,928 $187,286
7 0.5 6.0 $4,248,959 $2,564,088 0.020 $84,979 $51,282
8 2.0 6.0 $4,480,943 $2,647,596 0.020 $89,619 $52,952
9 5.0 6.0 $4,683,662 $2,699,728 0.020 $93,673 $53,995
TOTALS: 1.000 $2,545,246 $2,362,749
NOTES:
The Net Present Values presented above include $1,201,640 capital cost for GAC, $1,254,283 capital cost
for HiPOx, and annualized estimates for operations, maintenance and carbon/chemical expenses.
1 Assumed MTBE contamination
levels 2 Assumed TBA contamination
levels
3 Assumed probability of any one scenario occurring
19
1 2 3 4 5 6 7 8 9
HiPOx Option
GAC Option
$0
$500,000
$1,000,000
$1,500,000
$2,000,000
$2,500,000
$3,000,000
$3,500,000
$4,000,000
$4,500,000
$5,000,000
15 Year Annualized CostContamination Scenario
Figure 1 - NPV of GAC vs. HiPOX Treatment at Bakersfield Well
HiPOx Option GAC Option
References and Resources (cont’d.)
Tech Notes, Vol. 1, No. 10, MTBE, US Filter Westates Technical Department, Los Angeles, California, December 1999
Tech Notes, No. 4, US Filter Westates Technical Department, Los Angeles, California, August 1999
MTBE Mitigation, Pilot testing, Treatment and Related Litigation for the South Tahoe Public Utility District, Reitz, Redlin,
Zaninovich, & McAdams, 2003 (presented at the ASCE 2003 World Water & Environmental Resources Congress,
Philadelphia, PA, and at the AWWA 2003 Water Quality Technology Conference, Philadelphia, PA)
Pilot Plant Study Report for MTBE Removal Using Granular Activated Carbon, Tata Lane Well Water at Clement Well
Site, Boyle Engineering Corporation, Fresno, CA, December 2003
Acknowledgements
Special thanks to John A. Thiel, P.E./MBA for his expertise and contributions, guidance, and insight in preparing this
study. John is the Managing Engineer for for the South Tahoe Public Utility District, and may be reached at
530.543.6209, or e-mail: jthiel@stpud.dst.ca.us.
Nick Zaninovich was formerly the Senior Engineer with the South Tahoe Public Utility District. During his tenure, he
managed design, permitting, construction and commissioning of two advanced oxidation treatment facilities to destroy
MTBE contamination in drinking water for the District. Mr. Zaninovich is a Registered Professional Engineer in the states
of California and Nevada, and is licensed as a California Water Distribution System Operator, Level IV. He can currently
be reached at 530.577.8322, or e-mail nick1045@bigvalley.net. He is currently self-employed as a consulting civil
engineer.
20