WANG ET AL. | 107:5 • JOURNAL AWWA | MAY 2015 87

n California, where water has long been recognized as a scarce and

valuable resource, the Orange County (Calif.) Water District (OCWD)

has demonstrated commitment to advanced water purification and

reuse for more than three decades. Completion of the initial expansion

of OCWD’s groundwater replenishment system (GWRS) early in 2015

made the world’s largest advanced water purification system for potable reuse

even bigger.

OCWD’s original GWRS, completed in 2008, has an advanced water puri-

fication facility (AWPF) that purifies secondary effluent through microfiltra-

tion, reverse osmosis (RO), and advanced oxidation using ultraviolet (UV)

light and hydrogen peroxide. Purified water from the AWPF is injected into

groundwater aquifers to minimize seawater intrusion and replenish local

groundwater supplies.

The initial expansion of the GWRS provided an opportunity to enhance

AWPF processes while expanding production capacity from 70 to 100 mgd.

Enhancements focused on energy-recovery devices on the new RO units, sec-

ondary effluent flow equalization, and alternative treated water quality targets,

as well as new technologies to enhance the reliability and flexibility of post-

treatment operations. Evaluation and enhancement of post-treatment stabili-

zation processes at the AWPF were especially important for system reliability

and flexibility.

A RECENT EXPANSION OF

ORANGE COUNTY (CALIF.)

WATER DISTRICT’S

ADVANCED WATER

PURIFICATION FACILITY

ALLOWED THE

ENHANCEMENT OF

ENERGY-RECOVERY DEVICES

ON NEW REVERSE OSMOSIS

UNITS, SECONDARY

EFFLUENT FLOW

EQUALIZATION, AND

ALTERNATIVE TREATED

WATER QUALITY TARGETS.

SUNNY WANG, MEHUL PATEL, BILL DUNIVIN, JAMES CLARK,

STEVEN FOELLMI, AND CATHERINE (SPENCER) DIPIETRO

Post-treatment Optimization

to Enhance Groundwater

Recharge Operations

2015 © American Water Works Association

88 MAY 2015 | JOURNAL AWWA • 107:5 | WANG ET AL.

The post-treatment regime at the

AWPF consists of decarbonation

(carbon dioxide removal) and the

addition of lime to stabilize the

purified water or final product

water (FPW). However, over-

decarbonation of RO permeate

(ROP) often resulted in inadequate

carbonates—and thus inadequate

alkalinity—even after lime was

added. Inadequate carbonate alka-

linity made it difficult to control

finished water pH; therefore, it was

challenging to maintain ideal corro-

sion protection parameters and suit-

ability for groundwater recharge.

Improvements implemented as part

of the initial expansion supported

OCWD water quality objectives by

enhancing post-treatment operation

and reliability and producing more

consistent finished-water quality.

POST-TREATMENT OPTIMIZATION

WITH THE END GOAL IN MIND

Passing the entire feed flow

through the RO membranes at the

AWPF effectively removes contam-

inants—along with most of the dis-

solved constituents in the water.

The resulting ROP is highly corrosive

(low in dissolved solids, including

hardness and alkalinity as well as

pH) and requires post-treatment

stabilization. In such instances,

appropriate post-treatment is crit-

ical and must accomplish three

primary water quality objectives:

produce water that meets regula-

tory requirements, that is not cor-

rosive to the conveyance system

(e.g., to cement–mortar lining on

the GWRS conveyance piping), and

with low scaling potential in order

to prevent plugging at groundwa-

ter injection wells.

The project team adopted a com-

prehensive approach for the entire

post-treatment regime (from decar-

bonation through injection) to

increase the FPW’s alkalinity and

reliability to meet water quality

objectives. With this end goal in

mind, an alternative FPW goal was

developed for the GWRS that

enhanced reliability and consistency

of the FPW water quality for ground-

water injection. The original and

alternative post-treatment goals are

summarized in Table 1.

The benefits of changing the FPW

target were as follows:

• The potential for calcium car-

bonate (CaCO

)

scaling at a

seawater intrusion barrier would

be controlled by maintaining pH

below the values at which

CaCO

forms (e.g., <8.0).

• The increased alkalinity and

lower pH would increase buffer

intensity (often referred to as

buffer capacity). Buffer intensity

is a measure of parameters such

as inorganic carbonates or phos-

phates (if present) in the water

that resist pH change. A higher

buffer intensity could aid in sta-

bilizing the treated water pH.

• The increased alkalinity and cal-

cium levels would be more pro-

tective of the cement mortar on

the cement mortar–lined con-

veyance pipelines.

• The free carbon dioxide levels

would be higher compared with

the current average, ensuring

that sufficient carbon dioxide is

available for conversion to

bicarbonate when lime is added.

• The FPW would be more repre-

sentative of groundwater quality

from natural recharge (rainfall).

The alternative FPW goal would

result in more stable product water

(e.g., lower susceptibility to pH

variations) and provide better

protection for downstream convey-

ance pipelines. Research on ROP

post-treatment (AWWA Research

Foundation 1996) has demonstrated

that finished water with calcium

levels in excess of 20 mg/L as CaCO

alkalinity values in excess of 50 mg/L

as CaCO

and slightly positive

CaCO

precipitation potential values

was more protective of cement mor-

tar–lined pipe than water with lower

calcium and carbonate concentra-

tions. Higher alkalinity and calcium

levels correspond to lower rates of

cement–mortar degradation, even if

calcium carbonate precipitation

potential is slightly negative (AWWA

Research Foundation 1996). Thus, it

was recommended that OCWD use

a potential CaCO

precipitation tar-

get that was slightly negative (i.e.,

−5 to 0 mg/L) in combination with

TABLE 1 Summary of current and recommended post-treatment FPW goals

Parameter Original Goal Alternative FPW Goal

pH 8.5–8.8 7.6–7.9

Alkalinity—mg/L as CaCO

40 40–50

Calcium—mg/L 3–4 10–13

Hardness—mg/L as CaCO

7–10 25–33

Free carbon dioxide—mg/L None <3

Buffer intensity None 0.100

CCPP—mg/L as CaCO

−4 to −2 −3

CaCO

—calcium carbonate, CCPP—calcium carbonate precipitation potential, FPW—final product water

The post-treatment regime at the advanced water

purification facility consists of decarbonation (carbon

dioxide removal) and the addition of lime to stabilize

the purified water or final product water.

2015 © American Water Works Association

WANG ET AL. | 107:5 • JOURNAL AWWA | MAY 2015 89

the proposed finished water targets

for calcium, alkalinity, and pH to

protect against cement–mortar cor-

rosion. The project team recom-

mended and implemented additional

improvements to support the alterna-

tive FPW goal and increase opera-

tional flexibility and reliability for

the initial expansion of the GWRS,

including the following:

• A new bypass guideline for the

decarbonation towers, following

UV advanced oxidation process,

was developed to ensure that

adequate bicarbonate alkalinity

and buffering capacity were

maintained.

• A new lime slurry feed system

was pilot-tested and incorpo-

rated to improve the quality and

reliability of the entire lime feed

system from dry hydrated lime

feed through slurry makeup and

lime saturation.

DECARBONATION BYPASS

The AWPF’s current average

bypass percentage is approximately

20–30%, with 70–80% of the UV

product water (UVP) flow directed

through the decarbonators. The pri-

mary reason for sending the majority

of the UVP flow through the decar-

bonators is to minimize the costs of

chemicals that raise the pH of the

ROP or UVP. However, given the low

alkalinity values of the water after

RO and UV, coupled with low but

variable pH, the decarbonation sys-

tem must be operated to maintain

sufficient carbon dioxide that can be

converted to bicarbonate alkalinity

when lime is added.

New operating guidelines for the

amount of flow bypassing the decar-

bonators were developed to increase

alkalinity and support the alternative

FPW goal. These guidelines, shown

in Table 2, are used when the UVP

pH ranges from 4.9 to 5.8. If the

UVP pH is 5.9 or greater, as little

UVP as possible should be decarbon-

ated; if the UVP pH is below 4.8, all

of the water should be decarbonated.

LIME ADDITION

A side-by-side pilot test was con-

ducted to compare the performance

of the original flow-paced lime slurry

system against a weight-controlled

lime-slaking system.

This test was

essentially conducted at full scale

with the original lime slurry system

feeding one saturator—lime satura-

tor A—and a temporary weight-

controlled lime-slaking system feed-

ing another saturator—lime saturator

B. Each lime slurry makeup system

fed a separate lime saturator at the

AWPF. The photograph on page 91

shows the pilot-test facility.

On the basis of data collected from

the pilot test, the weight-controlled

lime-slaking system produced a more

consistent and repeatable lime slurry

than the original lime slurry system

(see percent solids data in Table 3)

with respect to the target percent

concentration selected (7%). The

standard deviation of the weight-

controlled lime-slaking system slurry

TABLE 2 UVP bypass/decarbonation flow percentages as a function of

UVP pH <5.9

pH of UVP

UVP Bypass

UVP Decarbonation

% Estimated pH of DPW

5.8

65 35 6.1

5.7

45 55 6.1

5.6

40 60 6.0

5.5

28 72 6.0

5.4 23 77 6.0

5.3 18 82 5.9

5.2 12 88 5.9

5.1 7 93 5.9

5.0 5 95 5.8

4.9 2 98 5.8

DPW—decarbonated product water, UVP—ultraviolet product water

Range of pH planned by Orange County Water District for future reverse osmosis operations

TABLE 3 Comparison of lime slurry percent solids data

Measure

Weight-Controlled Lime-Slaking System Original Lime Slurry System

Slurry Feed

gpm

Saturator Flow

gpm

Lime Slurry

% solids

Slurry Feed

gpm

Saturator Flow

gpm

Lime Slurry

% solids

Average 5.66 264 7.14 5.44 263 8.74

Minimum 4.25 176 7.05 3.75 184 7.24

Maximum 7.75 353 7.30 7.25 420 9.94

Standard deviation 0.91 46 0.08 0.88 43 0.69

Operation set point for both systems = 7% solids

2015 © American Water Works Association

90 MAY 2015 | JOURNAL AWWA • 107:5 | WANG ET AL.

was only 0.08 compared with 0.70

for the original lime slurry system.

The lower variability of the lime

slurry concentration produced by the

weight-controlled lime-slaking sys-

tem resulted in a more accurate lime

solution produced to stabilize the

finished product water.

Performance of the two saturators

with respect to percent solids, pH,

conductivity, total dissolved solids,

and alkalinity was comparable. How-

ever, turbidity of the saturated lime

solution from saturator B (lime slurry

fed by the weight-controlled lime-

slaking system) was two units lower

compared with saturator A (lime

slurry fed from the original lime

slurry system)—5.8 versus 3.5 ntu.

A plot of the two saturator effluent

turbidities over the course of the pilot

test is shown in Figure 1. The red-

shaded region indicates 1 standard

deviation of the lime saturator efflu-

ent (saturator A) turbidity, fed by the

existing flow-paced lime slurry sys-

tem. The blue-shaded region indicates

1 standard deviation of the lime satu-

rator effluent (saturator B) turbidity,

fed by the new weight-controlled lime

slurry system. The lower saturated

lime–effluent turbidity observed in

saturator B suggests that the new

weight-controlled lime-slaking system

produces a lime slurry that reacts

more readily and dissolves better than

the original flow-paced lime slurry

system. The lower saturated lime

effluent turbidity readings in lime

saturator B were also supported by

Secchi disk readings that were gener-

ally 1 ft clearer when compared with

lime saturator A. The lower turbidity

and better clarity in saturator B sug-

gest that fewer particulate/colloidal

impurities are produced with the

weight-controlled lime-slaking sys-

tem because the system’s slaker and

aging tank ensure adequate time for

the dry, hydrated lime to fully react

with the dilution water.

The effluent turbidity of the two

saturators is plotted against AWPF

production flow in Figure 2. The lime

saturator fed by the original lime slurry

system exhibited an increase in effluent

0.0

1. 0

2.0

3.0

4.0

5.0

6.0

7. 0

8.0

9.0

10.0

11.0

10/24/2009

11/03/2009

11/13/2009

11/23/2009

12/03/2009

12/13/2009

12/23/2009

01/02/2010

01/12/2010

Turbidity—ntu

Date

New weight-controlled lime slurry system

New weight-controlled lime slurry system average—3.5 ± 1. 7

Existing flow-paced lime slurry system

Existing flow-paced lime slurry system average—5.8 ± 2.5

FIGURE 1 Comparison of lime saturator efuent turbidity fed by existing

ow-paced lime slurry system versus weight-controlled lime

slurry system

12.0

0.00

2.00

4.00

6.00

8.00

10.00

12.00

45 50 55 60 65 70

Turbidity—ntu

AWPF Production Flow—mgd

FIGURE 2 Comparison of lime saturator efuent turbidity versus AWPF

efuent ow

New weight-controlled lime slurry system

Linear (new weight-controlled lime slurry system)

Existing flow-paced lime slurry system

Linear (existing flow-paced lime slurry system)

AWPF—advanced water purification facility

WANG ET AL. | 107:5 • JOURNAL AWWA | MAY 2015 91

turbidity with increasing AWPF pro-

duction when a greater amount of lime

was needed. The saturator fed by the

weight-controlled lime-slaking system

was not affected by varying lime

demand and AWPF production. Dur-

ing high-flow periods, more lime is

needed to stabilize the FPW, and batch

reaction time in the original lime slurry

system dramatically decreases because

it is a flow-paced system. Variable

AWPF production and lime demand

decrease overall reaction time and the

lime slurry consistency produced by

the original system during peak

demand periods. Because the weight-

controlled lime-slaking system oper-

ates in a true batch mode and accom-

modates varying lime demand through

the aging tank, consistent lime slurry is

produced within the weight-controlled

lime slaker regardless of lime demand

or AWPF production.

Pilot-test results (e.g., particle size

distribution, turbidity, percent sol-

ids) showed that the new weight-

controlled lime slurry makeup system

provides a more stable lime dose that

is not affected by diurnal flow varia-

tions at the AWPF. The project team

determined that the consistency and

accuracy of the lime slurry produced

by the weight-controlled lime-slaking

system would provide OCWD with

additional control over the post-

treatment process at the AWPF.

CONCLUSION

The initial expansion of OCWD’s

GWRS facilitated a holistic approach

to optimize post-treatment operations

at the AWPF and to enhance reliability

in producing water that meets regula-

tory requirements and will not corrode

the conveyance and groundwater injec-

tion facilities. The alternative FPW goal

and supporting process improvements

will give OCWD operation staff more

flexibility and reliability in meeting fin-

ished water quality targets to minimize

corrosion and enhance groundwater

recharge operations.

ENDNOTES

RDP Technologies, Norristown, Pa.

ZMI Portec Chemical Processing Group,

Sibley, Iowa

ABOUT THE AUTHORS

Sunny Wang (to

whom

correspondence

may be

addressed) is a

senior process

engineer in Black

& Veatch’s Water Technology

Group, 800 Wilshire Blvd., Ste.

600, Los Angeles, CA 90017

USA; wangsunny@bv.com.

He served as task leader for post-

treatment improvements for

initial expansion of Orange

County Water District’s

(OCWD’s) groundwater

replenishment system (GWRS)

and specializes in process

optimization and design of

The lime-slaking pilot unit consists of a slaker tank, slurry-aging tank, and transfer pumps feeding

lime saturator B during side-by-side testing. Photo courtesy of Orange County Water District.

Bulk bag loader

Transfer screw

Slurry-aging tank Slaker tank

Purified water from Orange County Water District’s (OCWD’s) Advanced Water Purification

Facility, highlighted here to distinguish it from the adjacent OCWD Wastewater Treatment

Plant, is injected into groundwater aquifers to minimize seawater intrusion and replenish

local groundwater supplies. Photo courtesy of Orange County Water District.

92 MAY 2015 | JOURNAL AWWA • 107:5 | WANG ET AL.

advanced water and wastewater

treatment technologies. Mehul Patel

is the GWRS program manager at the

OCWD, Fountain, Calif., and is the

former president of the American

Membrane Technology Association.

William Dunivin is the director of

water production at OCWD,

Fountain, Calif. James Clark is a

senior vice-president with Black &

Veatch, Los Angeles, Calif., and

served as the project manager for the

initial expansion of OCWD’s GWRS.

Steven Foellmi is a vice-president

with Black & Veatch, Irvine, Calif.,

and served as the engineering

manager for the initial expansion of

OCWD’s GWRS. Catherine

(Spencer) DiPietro, who participated

in the project as a process engineer

with Black & Veatch, is now the

president of Aquarius Engineering,

Pownal, Maine.

http://dx.doi.org/10.5942/jawwa.2015.107.0074

REFERENCE

AWWA Research Foundation, 1996 (2nd ed.).

Internal Corrosion of Water Distribution

Systems. AWWA Research Foundation,

Denver, Colo.