IMPROVING ENERGY EFFICIENCY OF LOW SALINITY DESALINATION
(WASTEWATER AND BRACKISH WATER) USING ENERGY RECOVERY
DEVICE – CASE STUDIES
David G. Argo (B&V, Irvine, CA)
Mehul Patel (OCWD, Fountain Valley, CA)
Shivaji Deshmukh (West Basin Municipal Water District, El Segundo, CA[Formerly With OCWD]
Srinivas (Vasu) Veerapaneni (B&V, Kansas City, MO)
Sunny Wang (B&V, Los Angeles, CA)
Overview
The Initial Expansion of the Groundwater Replenishment (GWR) System Project (Expansion Project) is
a proposed expansion of Orange County Water District’s (OCWD or District) Advanced Water
Purification Facility (AWPF) located in Fountain Valley, California. This paper describes the energy
recovery and reverse osmosis (RO) process optimization features of the Expansion Project.
The GWR System consists of three major components: (1) the AWPF and pumping stations; (2) a
transmission pipeline connecting the treatment facilities to existing recharge basins; and (3) an existing
seawater intrusion barrier. The expansion will add 30 million gallons per day (mgd) to the production
capacity of the AWPF. A major portion of the Expansion Project was to review existing design and
operations of the RO system at the AWPF and evaluate applicability of energy recovery devices (ERDs)
for the new RO units. Based on this evaluation, preliminary design criteria for the 30 mgd expansion
were developed to enhance the energy efficiency and operability of the new RO units.
Existing System
RO System Facilities
The existing RO system at the AWPF consists of RO transfer pumps, cartridge filters, RO feed pumps,
the RO units, RO Flush System, and RO Clean-In-Place (CIP) system. A summary of the design criteria
for the existing RO system is provided in Table 1.
Recommendations to Improve New RO Units for AWPF Expansion
As part of the AWPF expansion, operating data from the existing RO system was reviewed to evaluate
its performance and identify areas that could be enhanced. After careful review of the existing RO
system design and operating data, guidelines to improve overall RO performance were developed:
 Provide additional online instrumentation to improve performance monitoring capabilities of
each individual RO stage within each RO unit.
 Provide flux balancing capabilities between RO stages to improve overall membrane
performance and life.
 Minimize effects of concentrate polarization by maintaining a beta factor of less than 1.13 and
maintain adequate concentrate flow (~18 gallons per minute [gpm] per pressure vessel).
Table 1: Design Criteria of Existing RO System
Parameter Value
RO Transfer Pumps
No. of Pumps 4 duty, 1 standby
Design Capacity, each 23,000 gpm
Total Pumping Capacity (Duty) 92,000 gpm (132 mgd)
Rated Head 176 ft
Power, each 1,250 horsepower (hp)
Cartridge Filters
No. of Cartridge Filters 10 duty
Max capacity, each 11.75 mgd
Max Differential Pressure 15 pounds per square inch (psi)
RO Feed Pumps
No. of Pumps 14 duty, 1 standby
Design Capacity, each 4,340 gpm
Rated Head 681 ft
Power, each 1,000 hp
RO Units
No. of RO Units 14 duty, 1 standby
No. of RO Trains 5
No. of RO Units per Train 3
Design Production Capacity, each unit 5 mgd
Design Production Capacity, total 70 mgd
Number of Stages per RO Unit 3
RO System Recovery 85 percent
RO System Permeate Flux 12 gallons per square foot per
day (gfd)
RO Flush System
Flush Rate 1,000 gpm
Flush Duration 20 minutes
No. of RO Flush Pumps 3 duty
Pump Capacity, each 1,000 gpm
Rated Head 120 ft
Power, each 50 hp
Energy Recovery Device Implementation
Expansion of the AWPF provides OCWD an opportunity to enhance the performance of the RO system
by incorporating ERDs and implementing several process enhancements in the new RO units. The
project included a review of available ERDs on the market, an analysis of the performance of the
existing RO units, an assessment of ERD options for the District, and an energy recovery cost analysis.
Energy Recovery Devices
RO systems consume significantly more energy than other treatment processes at the AWPF. RO
membranes have a higher energy requirement in order to overcome the osmotic pressure of feed water
that is concentrated by a factor of approximately 6.7 within the RO system. Further, fouling of the
membranes by organic, inorganic, and biological constituents is inevitable for wastewater applications
and contributes to increased energy consumption over time. This results in a high amount of energy that
is wasted on the concentrate stream that can be recovered and used beneficially within the system.
As part of the project, a review of currently available ERDs was conducted: centrifugal force based
(Pelton Turbine, Turbo Energy, and Hydraulic Pressure Boosters) and positive displacement (work
exchanger and pressure exchanger).
Positive displacement ERDs can ideally be used to augment the high pressure pump feeding the 1
st
stage. Typically high pressure pumps pressurize feed flow that is equivalent to the permeate flow, and
the ERDs pressurize the flow equal to the concentrate flow. A booster pump is required to increase the
pressure of the ERD effluent to account for losses within the system. Use of these devices for inter-stage
boosting requires two pumps – one to convey the water to the ERD and another to further boost the
pressurized feed exiting the ERD device. Further, the potential for mixing of concentrate with feed flow
could impact permeate water quality. Hence, they are not ideally suitable for balancing the flux between
the various stages of a multi-stage RO system, such as the system at OCWD. In contrast, centrifugal
devices provide a smaller footprint, lower capital costs, and ability to balance fluxes between stages
while recovering energy. Although the efficiency of these devices in recovering energy is lower, these
devices are more appropriate for application at OCWD. Further, to allow flexibility for flux balancing
throughout membrane life, use of an ERD equipped with a motor is preferable. Such a device provides
the necessary increase in head for later stages as needed. Thus, the use of centrifugal ERDs was selected
for incorporation into the new RO units at the AWPF. These devices will be used to recover energy
from the concentrate. The recovered energy will be used to boost the feed pressure of the 2
nd
and 3
rd
stage. An evaluation was then conducted of the performance of existing RO unit without the ERD,
followed by use of ERD for boosting the 2
nd
and 3
rd
stages. Based on this evaluation, the final
configuration was developed and is discussed below.
Performance of Existing RO Units
The performance of existing RO units when treating feed water with average water quality was
evaluated. The fouling factor was adjusted to match the current performance observed. Typically, the
fouling rate during initial years of operation is high for wastewater applications and is expected to level
off. However, it is difficult to estimate the fouling rate To which the system will be stabilized. Hence,
two cases were developed to represent performance at the end of membrane life. In one case, the rate of
fouling would result in feed pressure of 225 psi at the end of membrane life. In the second case, the
current rate of fouling was expected to continue in a worst case scenario, resulting in feed pressure of
300 psi at the end of membrane life.
With new membranes (0 year life), the permeate flux of the 1
st
stage lead element is very high (20.8
gfd), and the lag elements of the 3
rd
stage make hardly any permeate. The flux is highly imbalanced
between the various stages, ranging from 17.2 gfd for the 1
st
stage to 8.7 gfd for the 2
nd
stage and 1.7 gfd
for the 3
rd
stage. In addition, the beta value for lag elements of the 1
st
stage is also higher than a
typically recommended value of 1.13. The recovery of these elements with high beta value is expected
to be greater than the desirable limit of 15 percent recovery for individual elements to minimize scaling
and fouling.
Due to significantly higher fluxes and element recovery, the 1
st
stage is expected to foul at a faster rate
than other stages. This results in the 1
st
stage making less permeate over time, and other stages making
more permeate by increasing the overall system operating pressure. The fluxes of individual stages
eventually balance to accepted ranges toward the end of membrane life. This is likely the intent of the
original design, but is achieved at the cost of higher operating expenses (energy required to overcome
fouling of 1
st
stage RO membranes). By balancing the fluxes from initial start-up, ERDs can minimize
the increase in energy consumption and reduce overall energy consumption.
Flux Balancing Through ERDs
Balancing the membrane fluxes of various stages involves reducing the high flux of 1
st
stage membranes
and increasing that of subsequent RO stages. Reducing the high 1
st
stage flux for the lead elements and
overall flux would result in a decrease in fouling rate. The membrane flux of other stages (2
nd
and 3
rd
)
would be increased to compensate for decreased permeate flow from 1
st
stage. This also would reduce
the individual element recovery in the 1
st
stage. In the existing RO units, the element recovery for some
of the elements is greater than the recommended value of 15 percent. An additional benefit would be
increased concentrate flow in the second stage, increasing the cross flow velocity and reducing the
fouling potential. Flux balancing could be achieved through three options: permeate throttling, inter-
stage booster pump, and an ERD integrated with a booster pump. After conducting a detailed analysis
looking at energy and process benefits, an ERD integrated with a booster pump was selected for
expansion of the AWPF. The ERD integrated with a booster pump will allow OCWD to reduce overall
energy consumption of the RO units as well as provide flux balancing capabilities, via motor and
booster pump, to prolong membrane life.
A schematic of RO operation with and without ERD is provided on Figures 1 and 2, respectively.
Further reduction in energy consumption per RO unit is possible by incorporating an ERD with a
booster pump. Preliminary RO membrane projections indicate that the inter-stage booster pump motor
will be operated initially (approximately first two years of operation) with external power supplementing
the ERD. It may not be necessary toward the end of the membrane life as concentrate pressure increases
as the membranes are fouled, thus increasing amount of energy recovered from the concentrate at higher
pressures. Since the rate of fouling to be expected in the future is not clear, having an ERD with a motor
will provide a high degree of flexibility for system operation.
FM
Permeate (12 psi)
Feed
Concentrate
5.88 mgd
FM
FM
C
C
C
FM
FM
CC
CC
CC
5 mgd
0.88 mgd
2.16 mgd
1 mgd
140 psi
112 psi
95 psi
3.76 mgd
1.16 mgd
0.11 mgd
73 psi
2.16 mgd
1 mgd
140 psi
112 psi
95 psi
3.76 mgd
1.16 mgd
0.11 mgd
73 psi
17.6 gfd 8.3 gfd 1.2 gfd17.6 gfd 8.3 gfd 1.2 gfd
Fluxes
Figure 1. Schematic of Existing RO Units Without ERD
FM
Permeate (13.5 psi)
Feed
Concentrate
Energy recovery device w/motorEnergy recovery device w/motor
5.88 mgd
1.58 mgd
FM
FM
CC
CC
CC
5 mgd
0.88 mgd
2.16 mgd
1 mgd
140 psi
112 psi
95 psi
3.76 mgd
1.16 mgd
0.11 mgd
73 psi
2.16 mgd
1 mgd
140 psi
112 psi
95 psi
3.76 mgd
1.16 mgd
0.11 mgd
73 psi
124 psi
2.7 mgd
1.12 mgd
3.23 mgd
0.23 mgd
149 psi
102 psi
82 psi
124 psi
2.7 mgd
1.12 mgd
3.23 mgd
0.23 mgd
149 psi
102 psi
82 psi
17.6 gfd 8.3 gfd 1.7 gfd17.6 gfd 8.3 gfd 1.7 gfd
14.4 gfd 12.2 gfd 3.8 gfd14.4 gfd 12.2 gfd 3.8 gfd
Fluxes
Figure 2. Schematic of New RO Units with ERD (Existing Values are Shown in Grey)
Energy Recovery Cost Analysis
The average energy consumption over the 5 year period for the existing RO system, based on feed
pressure increase from 183 psi to 224 psi, is approximately 481 kilowatts (kW). Given that the feed
pressure is only increasing by approximately 20 percent, use of a linear average is reasonable. The
estimated energy consumption of the existing RO system with an ERD integrated with a motor is
estimated to range from 415 kW to 504 kW. The average energy consumption is approximately 459.5
kW. The average energy consumption of the RO system with an ERD is approximately 21.5 kW less
than an RO without an ERD. The difference of 21.5 kW amounts to savings of approximately $17,000
per year at an energy cost of $0.10/kWh and assuming that the RO units operate only 90 percent of the
time. A cost/payback period analysis for incorporating ERDs for the RO units was developed, showing
a simple payback period of five to nine years depending on operating conditions (nine years being
worst-case). Incorporation of an ERD with a motor was recommended for the each RO units as it would
reduce energy consumption, while also allowing means of balancing the fluxes within the RO system to
provide process benefits. Similar benefits would also apply if ERDs were added to the existing RO
system, but available space for ERDs needs to be evaluated.
RO System Design For AWPF Expansion
Expanding the AWPF production capacity by another 30 mgd (total of 100 mgd production capacity)
would require six additional RO units to be added. Each RO unit for the AWPF expansion will maintain
the existing 5 mgd production capacity per RO unit. However, a seven pressure vessel high arrangement
was selected (rather than the existing six pressure vessel high design) due to space constraints of the site
for the RO building expansion. The RO system design criteria for the plant expansion are presented in
Table 2.
Table 2
New RO System Design Criteria
Parameter Value
RO Feed Pumps
No. of Pumps 6
Design Capacity, each 4,340 gpm
Rated Head 681 ft
Power, each 1,000 hp
RO Units
No. of RO Units 6 duty
No. of RO Trains 2
No. of RO Units per Train 3
Design Production Capacity, each unit 5 mgd
Design Production Capacity, total 30 mgd
RO Membrane
RO Membrane Element Diameter 8-inches
RO Membrane Element Length 40-inches
RO Membrane Area per Element 400
or 440 ft
2
RO Element per Pressure Vessel 7
Number of Stages per RO Unit 3
No. of Pressure Vessels per RO Unit 150
RO Array Arrangement 77:49:24
RO System Recovery 85%
RO System Flux Between 10.9 and 12.0 gfd
ERD
Type
Concentrate pressure driven turbine type
ERD integrated with a booster pump
Drive Direct by motor
Concentrate Influent Flow to ERD 618 gpm @ 79-96 psi
Feed Water Influent to ERD 1907-1946 gpm @ 121-142 psi
Pressure Boost 35 psi
The new RO units were designed with expanded instrumentation as compared to the existing system to
provide an accurate assessment of its performance. Specifically, permeate flows from the three stages
were separated to incorporate additional instrumentation to monitor the performance of individual stages
and to enable tracking inter-stage performance. The additional instruction for the new RO units includes:
online flow for 2
nd
and 3
rd
stages and total permeate of each RO unit. Permeate flow from the 1
st
stage
was calculated by subtracting 2
nd
and 3
rd
stage permeate flows from the total permeate flow. Permeate
conductivity monitoring will be provided for each RO stage of each RO unit. Sample taps will be
provided on the permeate and feed lines of each stage. The functionality and layout of these
improvements were enhanced by the use of 3-D design, as shown on Figure 3.
Figure 3. 3-D Design
Findings, Conclusions, and Recommendations
Expansion of the AWPF provided OCWD with an opportunity to optimize and enhance the existing RO
system, as well as incorporate design features to provide a more efficient and easy to operate RO
system. A major addition to the new RO units provided for the AWPF expansion will be the inclusion
of ERDs that will reduce overall energy requirements of the RO system. The ERDs will also potentially
prolong membrane life by balancing the flux between the three RO stages. With approximately 29 kW
of energy saved per RO unit, the ERD has a payback period of five to nine years (assuming an energy
cost of $0.10 per kWh and 90 percent operating time) depending on operating conditions.
