6 Water Research Foundation Guidelines for Hex Chrome_20170726

1 Water
Research
Foundation
advancing the science of water
Guidelines for Hexavalent Chromium
Treatment Testing
Web Report #4418
Subject Area: Water Quality
r

Guidelines for
Hexavalent Chromium
Treatment Testing
Prepared by:
Nicole K. Blute, Ying Wu, and Brent Alspach
ARCADIS U.S., Inc., 888 West 6th Street, 3rd Floor, Los Angeles, California 90017
Contributors:
Chad Seidel
Jacobs Engineering
Phil Brandhuber
HDR Engineering
and
Issam Najm
Water Quality & Treatment Solutions
Sponsored by:
Water Research Foundation
6666 West Quincy Avenue, Denver, CO 80235-3098
Published by:
Water
Research
Foundation=µ
©2012 Water Research Foundation. ALL RIGHTS RESERVED.

DISCLAIMER
This study was funded by the Water Research Foundation (Foundation). The Foundation assumes no
responsibility for the content of the research study reported in this publication or for the opinions or
statements of fact expressed in the report. The mention of trade names for commercial products does not
represent or imply the approval or endorsement of the Foundation. This report is presented solely for
informational purposes.
Copyright © 2012
by Water Research Foundation
ALL RIGHTS RESERVED.
No part of this publication may be copied, reproduced
or otherwise utilized without permission.
©2012 Water Research Foundation. ALL RIGHTS RESERVED.

Introduction
Guidelines Development Approach
Benefit to Subscribers
How to Use this Document
Overview of Chromium Treatment Options
Pilot Testing Guidelines
Guidelines for Pilot -Scale Evaluation of Anion Exchange
System Configuration/Setup
System Operation
Monitoring
Guidelines for Pilot Evaluation of Reduction/Coagulation/Filtration
System Configuration/Setup
System Operation
Monitoring
Guidelines for Pilot Evaluation of High -Pressure Membranes (RO and NF)
System Configuration/Setup
System Operation
Monitoring
Reporting
References
1
1
1
1
2
3
5
5
8
8
10
10
12
13
14
14
16
17
18
19
Table of Contents
©2012 Water Research Foundation. ALL RIGHTS RESERVED.

Guidelines for Cr(VI) Treatment Testing
Introduction
The intent of these guidelines is to identify items that should be considered when
designing, implementing, and reporting on pilot testing of Cr(VI) treatment for drinking
water sources. The document is written to provide a common baseline of information to
consider while planning pilot testing for Cr(VI) removal. This document is intended to
establish sufficient commonalities between projects to enable comparison of future
testing results from Cr(VI) treatment pilot studies performed at different locations and
by different utilities, which will likely support technology costing for future regulatory
development.
Guidelines Development Approach
Malcolm-Pirnie/ARCADIS drafted a framework for the guidelines based on the testing
protocol used in the Cr(VI) treatment studies conducted at the City of Glendale,
California. Phil Brandhuber of HDR Engineering, Issam Najm of Water Quality &
Treatment Solutions, and Chad Seidel of Jacobs Engineering also contributed in
developing these guidelines through participation at workshops and as reviewers.
Benefit to Subscribers
Regulatory agencies need to have adequate information in order to establish a
maximum contaminant level (MCL) for Cr(VI) treatment. As few studies are available
on treatment technologies for Cr(VI) removal from drinking water to low levels (e.g.,
sub parts -per billion (µg/L) to single digit ppb range), consensus from a recent Oct.
2011 WaterRF-sponsored workshop on Cr(VI) was that more studies are necessary to
establish treatment technology applicability to other utilities and costs. Following these
guidelines should yield sufficient information to enable comparison of future pilot
testing results of currently known Cr(VI) treatment technologies so that regulators (e.g.,
USEPA, California Department of Public Health (CDPH), and other states) and utilities
can determine overall costs of the various treatment options and applicability of
technologies.
How to Use this Document
This document is intended to be used as a reference during pilot test planning. An
example flowchart of the process that could be used to incorporate these guidelines is
shown in Figure 1.
©2012 Water Research Foundation. ALL RIGHTS RESERVED.

Guidelines for Cr(VI) Treatment Testing
i
a-
Lj
Figure 1. Example Use of these Guidelines in a Pilot Testing Program
As a general recommendation, it is suggested that utilities work closely with their local
permitting agency (e.g., In California, CDPH) to ensure that the designed pilot testing
approach answers questions that might otherwise slow down the permitting of a
demonstration or full-scale system.
Overview of Chromium Treatment Options
Chromium exists in drinking water sources in two oxidation states: hexavalent
chromium, Cr(VI), and trivalent chromium, Cr(III). Current understanding is that Cr(VI)
represents a more significant health risk than Cr(III) in drinking water. Removal
technologies can be classified as predominantly removing either Cr(VI) or Cr(III).
Although reduction of Cr(VI) to Cr(III) can be accomplished using a variety of reducing
chemicals, Cr(III) must be removed from the water to avoid reoxidation of Cr(III) to
Cr(VI) in distribution systems, which has been shown for typical chlorine and
chloramine concentrations and distribution system residence times. The rapid
conversion between the two oxidation states also underlies the need to measure both
Cr(VI) and total Cr in pilot tests to determine Cr(VI) removal efficiency and whether all
Cr(III) available for reoxidation has been removed.
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 2

Guidelines for Cr(VI) Treatment Testing
Cr(VI) treatment options for drinking water have been studied over the past decade,
primarily in a research program overseen by the City of Glendale, California, involving
many parties (Blute and Kavounas, 2011). With the announced intention of California
Department of Public Health (CDPH) to propose a Cr(VI) MCL in 2013, new
technologies are coming to market. This document is focused on guidelines for testing
the three technology classes considered to be the most suitable for Cr(VI) removal at
this time. It should be noted that other promising technologies may emerge in the
future. The basic guidelines for these three technologies can be used as a reference to
determine testing strategies for evaluating other emerging technologies.
The three technologies for which guidelines are provided in this document include:
• Anion exchange (including single pass, non-regenerable Weak Base Anion
(WBA) exchange or regenerable Strong Base Anion (SBA) exchange),
• Reduction/coagulation/filtration (RCF), and
• High-pressure membrane filtration (e.g., reverse osmosis and nanofiltration).
II. Pilot Testing Guidelines
In light of the uncertainty associated with the potential Cr(VI) MCL, utilities should use
the lowest possible detection limits for Cr(VI) and total Cr. Currently, detection limits of
0.02 to 0.06 ppb is possible for Cr(VI), and approximately 0.1 ppb for total Cr.
Collection of data to these detection limits will allow for evaluation of treatment
effectiveness at low levels and maximize the value of the research in informing the
regulatory process. In addition to low detection limits for chromium, pilot testing should
strive to build a statistically robust dataset that can validate the treatment process and
achievement of goals. A focus on quality assurance and quality control goals is
necessary to establish trust in the dataset, and methods consistent with Standard
Methods (2005) are generally optimal.
A summary of key issues associated with each of the three technology types in these
guidelines is provided in Table 1. The issues can be divided into several categories,
including: Configuration/Setup; System Operations; Monitoring; and Reporting. A
technology -specific discussion follows the table and provides additional details about
the table entries.
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 3

Guidelines for Cr(VI) Treatment Testing
Table 1. Summary of Pilot Testing Guidelines for Leading Technologies
Anion Exchange
Pilot System • Column dimensions (for both SBA and WBA):
Resin depth to bed diameter >4:1
Configuration/ Column diameter to bead diameter (D/d) >26:1
Setup • Particle filtration to avoid clogging
• pH adjustment for WBA resins to 5.5 to 6.0
• Brine regeneration for SBA
• Brine treatment alternatives test for SBA
• Mid -depth sampling ports may be useful for WBA
Pilot System Key operating parameters:
• HLR (e.g., 5 — 15 gpm/sf)
Operation • Service flow rate (e.g., 1 — 5 gpm/cf)
• EBCT (e.g., 2 - 4 minutes)
• Bed life (e.g. bed volumes to breakthrough) for both
WBA and SBA
• Brine regeneration requirements (e.g., brine
concentration and regen. frequency for SBA)
Pilot Startup:
• Compounds that may leach from resins, including VOCs,
Monitoring SOCs, formaldehyde/ ketones, tentatively identified
compounds, nitrosamines
Routine:
• Cr(VI) and total Cr, competing anions (nitrate, sulfate,
phosphate, bicarbonate), physical parameters (pH,
temperature, conductivity, alkalinity), compounds that
leach, other heavy metals if present in raw water
• Sampling of raw water, after pH change, and effluent.
Consider mid -depth sampling.
• Sufficiently frequent Cr(VI) and total Cr sampling to
capture the breakthrough curves for both SBA and WBA
resins.
• Minimum number of 5 regeneration and operation cycles
to prove consistency/repeatability of SBA treatment
Residuals:
• Hazardous waste testing of resin (WBA) or brine (SBA);
uranium loading for WBA resin
Reduction/ Coagulation/ Filtration
Components:
Reduction (e.g., 30-45 minutes for a tank
approach)
Aeration/Coagulation (i.e., oxidation with air)
Rapid mix with polymer
Filtration approach (e.g., granular or MF)
pH adjustment if source water is greater than 7.7
to maximize Cr(VI) reduction
Key operating parameters:
• Iron -to -chromium mass ratio (e.g., 25:1-50:1
depending on influent Cr(VI) concentration
• HLR (e.g., 3-4 gpm/sf)
• Backwashing frequency and duration
Routine:
• Cr(VI) and total Cr, total iron, ferrous iron,
dissolved oxygen, silica, physical parameters
(pH, temperature, conductivity, turbidity, and
alkalinity).
• Sampling between each unit process.
• Sufficiently frequent Cr(VI) and total Cr sampling
to capture the reliability of the process, especially
particle breakthrough with granular media filters.
• Minimum number of 5 granular media each
experimental condition or several MF clean -in -
place cycles.
Residuals:
• Hazardous waste testing of dewatered solids;
testing of backwash water to ensure suitability for
recycle or disposal.
High -Pressure Membrane Filtration
• Element manufacturer models to determine initial
estimates of flux, pressures, recovery, and need for
antiscalants and/or acid
• Cartridge filters with pore sizes in the range of 3-5
µm are standard prefiltration equipment
• For feed waters high in particulates, additional
prefiltration (e.g., MF/UF) may be needed to
achieve recommended SDI of less than 3-5
• Solubility indices for membrane foulants
Key operating parameters:
• Flow (or flux)
• Feed, concentrate, and permeate pressures
• Recovery
Routine:
Cr(VI) and total Cr, dissolved physical parameters
(pH, temperature, conductivity, TDS, turbidity,
ORP), SDI.
Sampling of raw water, pretreated water, permeate,
concentrate.
Operate long enough for demonstrating cleaning
frequency at desired/optimal operating conditions
for full-scale.
Residuals:
• Hazardous waste testing of concentrate.
• Permeability trends and flux decline
Pilot Test • Testing objectives
• System operations
Reporting • Unintended consequences observed
• Residuals characterization and volumes generated
Post -treatment considerations
• Treatment technology integration into an existing treatment train
©2012 Water Research Foundation. ALL RIGHTS RESERVED.

Guidelines for Cr(VI) Treatment Testing
Guidelines for Pilot -Scale Evaluation of Anion Exchange
Two types of anion exchange resins have been demonstrated for Cr(VI) treatment in
drinking water: SBA resins that are regenerated with a salt brine and WBA resins that
are operated as single -pass without regeneration. Anion exchange resin works by
exchanging Cr(VI) in the form of chromate with another less -selective anion (e.g.,
chloride). As the resin exchange sites reach capacity for Cr(VI), the resin will need
regeneration (regenerable) or replacement (single -pass). For Cr(VI), single -pass resins
shown to be effective (with a capacity of more than fifty times that of regenerable SBA
resins) are WBA resins (McGuire et al., 2007; McGuire et al., 2006). The mechanism of
the higher WBA capacity involves reduction of Cr(VI) to Cr(lII) by the resin, but the
exact component of the resin performing the reduction is unknown.
System Configuration/Setup
Pilot testing columns should be equipped with flow control, columns that allow resin
loading and replacement, water sampling ports, effluent discharge piping, and
arrangements to allow resin backwashing and regeneration (if needed). The raw water
should be filtered upstream of the resin columns to remove large particles from the
feed water to prevent bed clogging and the need for backwashing. Backwashing may
disturb the mass transfer zone within the resin bed, reducing Cr(VI) removal
effectiveness.
Figure 2 shows an example schematic for a pilot testing configuration of WBA resins.
WBA resins currently on the market require pH adjustment to between 5.5 and 6.0,
which can be achieved with acid addition (e.g., hydrochloric acid or carbon dioxide).
Multiple columns can be included in the skid to allow side -by -side testing of multiple
resins or the same resin under different conditions. Caution must be taken to ensure
thorough mixing of the acid or carbon dioxide into the feed water in the WBA columns.
Adequacy of mixing can be checked by obtaining in -line pH measurements or using a
sampling device to minimize offgassing during sampling.
Multiple sampling ports can also be useful to sample from different layers of the resin
bed, such as 50% of the resin bed depth, to observe breakthrough earlier than at the
end of the resin bed depth. If the sample ports are inserted into the resin bed, a screen
on the inside of the sample ports is recommended to minimize resin accumulation
inside the sample ports. Other necessary sampling ports are column feed water and
effluent points.
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 5

Guidelines for Cr(VI) Treatment Testing
Raw water Raw
Filter
Mixing pH
Water
Adjusted
Sample
Influent
Tap
Sample
Acid
Tap
WBA
Resin
50
Sampling
Port
Flow meter
and totalizer
Effluent
Sample Post -
Tap Treatment
(full-scale)
To waste
Figure 2. Schematic of an Example Pilot Testing Configuration for WBA Resin
SBA resins will not likely require pH adjustment unless calcium carbonate precipitation
potential is significant and bed plugging is anticipated. The other major difference with
SBA resins compared with WBA resins is that the configuration should include brine
storage tanks (for fresh and spent brine) in addition to other similar components, as
shown in Figure 3.
Several factors influence the minimum column diameter and bed depth for pilot
columns of both SBA and WBA, including (1) aspect ratio of resin bed depth to column
diameter, and (2) ratio of column diameter to resin bead diameter (D/d) to minimize
wall effects. Dow recommends a minimum bed diameter of approximately 0.75 inches
and an aspect ratio of at least 4 (Dow, 2011). Most literature reports demonstrate that a
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 6

Guidelines for Cr(VI) Treatment Testing
ratio of 50 (AWWA, 1982) to 100 (Kawamura, 2000) is effective in avoiding wall effects
in filtration pilot columns. However, two other tests indicated that wall effects may not
be observed down to 26:1 (McLellan 2011; Lang et al., 1993). For testing in Glendale,
California of WBA and SBA resins, use of a 2.5 inch column diameter (a column
diameter to resin bead diameter of approximately 40) was proven to be effective at
representing full-scale bed life.
Post -treatment processes may also be necessary for the WBA process for corrosion
control. Air stripping of carbon dioxide gas or addition of base could increase the
effluent water pH and decrease corrosivity toward distribution materials.
Q
Raw water
Filter Raw
Spent
Water
Brine
Sample
Sample
Tap
Tap
Spent Brine and
Spent Brine
Treatment Process
SBA
Raw
I Possible
Water
Recycle for
Sample
I Multiple
Tap
Regeneration
Cycles?
Row meter
and totalizer
Effluent
Brine
Sample
Tap
To waste
Figure 3. Schematic of an Example Pilot Testing Configuration for SBA Resin
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 7

Guidelines for Cr(VI) Treatment Testing
System Operation
Key operating parameters for anion exchange column testing include:
• Hydraulic loading rate (HLR)
• Service flow rate
• Empty bed contact time (EBCT)
• Flow rate (depending on HLR and column diameter)
• Bed life (number of bed volumes of water treated up to a target goal)
• For SBA, brine regeneration conditions (brine concentration) and frequency
HLR and EBCT are the primary design parameters for WBA resin treatment, since both
impact facility footprint. As a rule of thumb, a typical EBCT for ion exchange resins is 2
to 3 minutes, a HLR of 5 to15 gpm per square foot, and a service flow rate of 1 to 5
gpm per cubic foot. Pilot testing should ideally use the same HLR and EBCT values for
the desired full-scale operations, and vendors of the media selected should be
contacted to ensure that parameters selected fall within the acceptable range for the
resin.
Monitoring
Startup Sampling
Pilot testing start-up will simulate resin loading/replacement at full-scale. Special water
quality monitoring is needed during column test start-up to ensure that other
contaminants are not introduced into the water from the resin. For example,
nitrosamines and known potential resin constituents (e.g., formaldehyde for phenol -
formaldehyde backbone resins) could be released from the resin. Ideally, a broad
scan of constituents including volatile organic compounds (VOCs) and synthetic
organic compounds (SOCs), as well as aldehydes/ketones and tentatively identified
compounds (TICs) is recommended to ensure that leaching of unexpected constituents
of health concern is not observed.
Routine Monitoring
Water quality parameters to be measured in anion exchange pilot testing should
include at a minimum 1) Cr(VI) and total chromium; 2) other contaminants that can be
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 8

Guidelines for Cr(VI) Treatment Testing
removed by the resin or that can potentially affect resin capacity for Cr(VI), such as
nitrate, sulfate, phosphate, and bicarbonate; 3) general physical parameters, such as
pH, temperature, conductivity and alkalinity; 4) potential contaminants released from
the resin, such as nitrosamines or aldehydes/ketones, if shown to be present in the
startup testing; and 5) heavy metals if present in raw water. SBA testing should also
include sampling of brine for regulated constituents to evaluate disposal options.
The primary sampling locations should include column influent and effluent, as well as
the raw water before pH reduction for WBA. Additionally, water from different sampling
depths in the resin bed could also be measured to obtain advance data of when to
expect breakthrough from the effluent sampling point. Samples should be collected at
a frequency high enough to capture the breakthrough curve and allow for identification
of bed usage rates at different desired effluent concentrations.
Operating conditions should be monitored to ensure stable testing conditions during
the test period. The operating parameters that should be checked and recorded
include feed water pressure, water flow rate for individual columns, and total water
volumes processed by individual columns. For SBA, monitoring of the brine
regeneration process should be added.
The resin columns should be visually inspected for color change, air bubbles, and any
abnormal appearance (note that use of transparent columns is ideal to enable viewing
of the columns during pilot testing). Resin color may become darker gradually over
time due to accumulation of chromium or natural organic matter on the resin.
However, algae growth on the resin, especially when resin is exposed to the sunlight in
summer, may also result in resin color change so the columns should be covered to
prevent long-term exposure to light through clear columns.
Contaminant Release During Operation
The impact of system shutdowns/startups should be evaluated to see whether
significant levels of contaminants are released due to equilibration of resin with the
pore water during a shutdown. Contaminants to be monitored include 1) contaminants
that can be removed by the resin, such as chromium and nitrate, and 2) contaminants
that can be released from the resin, such as components that make up the resin
material structure (e.g., formaldehyde components or nitroso- compounds).
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 9

Guidelines for Cr(VI) Treatment Testing
Residuals Characterization
Spent resin at the end of the study should be tested for waste characteristics using the
Toxicity Characteristic Leaching Procedure (TCLP), as well the California Waste
Extraction Test (WET) if the resin is intended to be disposed in California. Brine waste
must be similarly characterized (albeit without digestion of solids) using the limits in the
above methods for the liquid component. Treatment of spent brine may be needed to
prevent hazardous classification. Radioactive elements (such as uranium) on WBA
resins should be tested since the resins may accumulate radionuclides.
Test Period
The WBA column test should run until a target effluent chromium target level is
reached (at least 50% of the influent concentration). A complete breakthrough curve is
desired, which means chromium levels in resin effluent reach the feed water chromium
concentration, particularly if concentrations in the influent are low. Note that some high
capacity WBA resins may require a very long time (e.g., nine months to over a year) to
reach chromium saturation (e.g., on the order of 100,000 bed volumes). Sampling
along various media depths in the resin bed could be particularly useful for WBA resin
testing to anticipate length of testing halfway through.
Regenerable SBA resins will reach breakthrough much faster (on the order of several
thousand bed volumes) compared with WBA resins, requiring more frequent sampling
to characterize the breakthrough profile. A minimum of five regeneration cycles is
recommended to characterize breakthrough curves, and more regenerations may be
desired especially if brine recycle is investigated.
Guidelines for Pilot Evaluation of Reduction/Coagulation/Filtration
System Configuration/Setup
The RCF process consists of three processes: 1) reduction of Cr(VI) to Cr(III) using
ferrous sulfate or ferrous chloride, 2) coagulation, and 3) filtration. The coagulation
step is likely to include aeration to ensure that excess ferrous iron is fully oxidized to
ferric iron so that the iron will be removed with filtration. Polymer addition may also be
useful to enhance formation of large particles for granular media filtration. Figure 4
shows an example schematic of a pilot testing configuration for the RCF process.
Pilot testing of the RCF process can be performed by scaling down the full-scale
treatment processes with respect to flow rate. Other parameters, such as chemical
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 10

Guidelines for Cr(VI) Treatment Testing
doses, reduction time, aeration time, and filter HLR, should directly reflect the full-scale
process conditions. In Glendale, California, two 2-gpm pilot tests of the RCF process
were able to simulate the performance of a full-scale 100-gpm facility (Qin et al., 2005;
Malcolm Pirnie, 2008). Jar testing of the RCF process, including reduction and
aeration in the jars followed by filtration, can also provide information on key variables
such as the need for pH adjustment or the potential impact of water quality on process
effectiveness prior to design of pilot testing. Jar testing can provide a means of
establishing operating conditions to test at pilot -scale and assessing the potential
effectiveness of the process under different water quality conditions.
Sufficient time is required to enable full Cr(VI) reduction to Cr(III). For mixed tanks, the
reduction time found to be effective at Glendale was between 30 to 45 minutes
depending on the influent Cr(VI) concentration and iron dose. Aeration was found to be
helpful in ensuring full oxidation of iron, particularly at a higher Fe:Cr(VI) ratio that was
necessary for a lower influent Cr(VI) concentration. These variables can be
investigated in pilot testing for different water quality conditions.
Two potential approaches to filtration are possible, including granular media filtration
and micro- or ultrafiltration (MF/UF). Preliminary jar testing indicates that the iron floc
produced in the RCF process may be amendable to MF/UF and might remove more
significant quantities of total Cr than granular filtration.
A critical parameter in the RCF process is pH adjustment, due to the sensitivity of the
reduction reaction and ferric iron precipitation to pH. Jar testing performed for the
Association of California Water Agencies on four utilities' water qualities demonstrated
that pH reduction may be necessary if influent pH is higher than approximately 7.7.
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 11

Guidelines for Cr(VI) Treatment Testing
Reduction
Tank
Influent
Flow meter
and totalizer
Sample Tap
Spent backwash
Pump
water treatment
Raw water j Q
=
Raw
Backwash
Water
Sample
Sample
Tap
Tap
pH adjustment
Ferrous
Filtration
if necessary
iron
(Granular
media or
Microfiltration)
ReductionAera$oR
-
;
Rapid Mix with
- - -
polymer addition
- ' -
(if Granular
-
Filtration)
hil
Effluent
Reduction
Raw
Sample
Tank Effluent
Water
Tap
Sample Tap
Sample
Backwash
Tap
To waste
Figure 4. Schematic of an Example Pilot Testing Configuration for the RCF
Process
System Operation
Key operating parameters for RCF pilot testing include:
• Iron -to -chromium mass ratio
• Hydraulic loading rate (HLR) of filters
• Backwashing frequency and duration
The iron -to -chromium mass ratio is important in ensuring that sufficient reductant is
available to reduce the Cr(VI) while simultaneously building up large enough particles
for removal by filtration.
Hydraulic loading rates found to be effective in past testing of granular media filtration
in Glendale, California were 3 to 4 gpm/sf, whereas 6 gpm/sf resulted in breakthrough
of iron and chromium. HLR rates are important in design because they impact vessel
sizes, and backwashing frequency impacts the sizing of backwash water tanks and
dewatering processes. Improvements in the HLR, such as could be tested in piloting,
could reduce process cost and footprint.
The backwashing frequency and duration can be tested to determine the length of time
to develop headloss that necessitates backwashing and the backwashing procedure
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 12

Guidelines for Cr(VI) Treatment Testing
that effectively cleans the media. Backwash frequency and duration impact the
residuals volumes and disposal options and costs.
Monitoring
Routine Monitoring
No special startup monitoring is anticipated for RCF, except that additional samples
may be required to ensure process stability. Water quality parameters to be tested for
the RCF process should include at a minimum 1) Cr(VI) and total chromium; 2) ferrous
iron and total iron, 3) constituents that can impact the RCF process, including dissolved
oxygen and silica; and 4) general physical parameters, such as pH, temperature,
conductivity, turbidity, and alkalinity. Sampling locations are recommended after each
process in the treatment train to enable focus on individual process effectiveness for
optimization. Samples should be collected with sufficient frequency to ensure process
stability with respect to ferrous dosing and pH and to obtain enough data points to
evaluate the stability of the system in terms of Cr(VI) and total Cr removal.
Operating conditions should be monitored to ensure consistent testing conditions
during the test period. The operating parameters that should be checked and recorded
include water flow rate, ferrous dosing, pH adjustment chemical dosing, polymer
dosing, reduction tank mixer operation, filtration process backwashing routine
completion, and the presence of bubbles in the aeration tank.
Residuals Characterization
Dewatered solids collected from the RCF process should be tested for waste
characteristics using TCLP, as well as the California WET method for disposal in
California. Supernatant from backwash settling (i.e., spent backwash water) should be
analyzed for suitability for recycle or disposal.
Test Period
RCF testing can be significantly shorter in duration than WBA pilot testing, although
more components are necessary for the RCF configuration. An RCF system with
granular media filtration could effectively demonstrate the technology effectiveness
within several weeks to a month, representing at least 5 backwashing cycles (as
recommended by CDPH for pilot testing of coagulation/filtration processes for arsenic
removal). Additional runs may be desired to optimize the process. Individual states
may have other specific guidance on filtration testing duration.
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 13

Guidelines for Cr(VI) Treatment Testing
Demonstration that the system operating conditions (e.g., reduction time, filtration and
backwashing process, iron -to -chromium ratio, pH conditions, and polymer selected and
dose) achieve targeted Cr(VI) and total Cr goals can be short if proof of concept or
longer if system optimization for design is desired. If microfiltration is selected for
testing rather than gravity media filtration, longer timeframes for testing (e.g., several
months) may be necessary to identify design criteria for the microfiltration system.
Guidelines for Pilot Evaluation of High -Pressure Membranes (RO and NF)
Reverse osmosis (RO) and nanofiltration (NF) membranes reject water constituents on
the basis of molecular size/weight and charge characteristics. Larger molecules or
ions that carry a higher charge (positive or negative) and/or are more highly branched
will be rejected more efficiently. The converse is also true: smaller, less charged,
and/or more compact molecules will be rejected less efficiently. Because the two
primary Cr(VI) species in water — chromate (Cr04 2) and dichromate (Cr207- 2) — are
both larger, multivalent ions, Cr(VI) should be efficiently rejected by RO membranes
similar to sulfate. Testing of various RO and NF membranes at the bench scale
showed that effective Cr(VI) may be possible with these technologies. A recent study
by Rad et al. (2009) reported Cr(VI) rejection between 99.5 and 99.8 percent for RO
membranes. Cr(III), present primarily as the uncharged species Cr(OH)3, would be
rejected as a precipitate in its particulate form, although the soluble fraction would
generally not be as efficiently removed.
NF is similar to RO, with membranes that are less selective, particularly for monovalent
ions. Some NF membranes should be able to achieve comparable Cr(VI) removal at
lower operating cost. However, because there is no standard definition of
nanofiltration, different NF membrane products may exhibit varying rejection
characteristics, as observed by Brandhuber et al. (2004).
In addition to confirming the rejection of Cr(VI) under application -specific conditions,
high-pressure membrane testing is typically conducted to collect information regarding
system operations, including pretreatment refinement, chemical use, membrane fouling
rates, fluxes, recovery, and power consumption, all of which are important factors for
determining life cycle cost estimates.
System Configuration/Setup
Due to the importance of using pilot testing to collect representative operational data,
the RO or NF pilot system should duplicate full-scale conditions to the extent possible.
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 14

Guidelines for Cr(VI) Treatment Testing
Consequently, the system should consist of at least a single pressure vessel housing 6
to 8 RO elements, as is common full-scale design practice. A second stage consisting
of an additional pressure vessel of similar design (i.e., 6 to 8 elements, as was used in
the first stage) can be used to increase system recovery and should be pilot tested if
intended for use at full-scale, since the second stage will experience the highest
concentration of foulants. Either 8-inch (diameter) or 4-inch (diameter) elements can
be used. Prior to pilot system design, an element manufacturer's software can be
used to model different system configurations using its respective RO products,
yielding estimates of rejection, pressures, fluxes, recovery, etc.
Effective pretreatment is extremely important for RO and NF operation. Because
elements cannot be backwashed, the removal of particulate matter is critical. Cartridge
filters are standard pretreatment equipment on high-pressure membrane systems, and
pore sizes of 3 to 5 µm or smaller are recommended. If the source water contains high
levels of particulate matter, as is typical in surface water applications, membrane
filtration (i.e., MF or UF) may be necessary for pretreatment. A maximum silt density
index (SDI) of 5 is typically cited for RO feed water, although SDI values of 3 or lower
are often recommended.
Because RO and NF systems concentrate sparingly soluble salts that can scale
membranes and reduce efficiency, chemical pretreatment is also common (e.g., acid or
scale inhibitor). For example, acid may added to lower the pH and increase solubility
of some compounds. Acid dosing (e.g., sulfuric or hydrochloric acid) can be estimated
using any element manufacturer's modeling software. If the concentrations of scalants
such as silica, barium, strontium, and others are significant, such that acid cannot be
sufficiently effective to control scaling potential and/or the required doses are too high,
a proprietary antiscalant (i.e., scale inhibitor) may also be needed. Consequently, the
availability of a thorough range of representative water quality data is imperative for RO
and NF pilot system and pretreatment design.
Post -treatment processes may also be necessary, depending on the permeate water
quality and blending strategy. For example, if acid is used as pretreatment to control
scale, carbonate species may be converted to carbon dioxide gas, which is poorly
rejected by RO and NF membranes. In this case, air stripping can be used to removed
dissolved gases in permeate. In addition, because permeate is very low in dissolved
solids, the water is often corrosive. This corrosivity is addressed either by blending
with other water supplies prior to distribution whenever possible or via the addition of
alkalinity and calcium when necessary.
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 15

Guidelines for Cr(VI) Treatment Testing
High -Pressure
Membranes
g
r%Q4-
7 To waste
Raw water
Raw
Influent
Post-
Effluent
Water
Filter
Water
Treatment
Water
Sample
Sample
Concentrate (full-scale)
Sample
Tap
Acid Scale Tap
Tap
(if needed) Inhibitor
(if needed)
Figure 5. Schematic of an Example Pilot Testing Configuration for High -Pressure
Membranes
System Operation
Initial system operational parameters for high-pressure membrane pilot systems,
including flow (or flux), recovery, and pressures, are typically determined by the RO
element manufacturers' software. The element manufacturers should be consulted to
determine threshold performance benchmarks beyond which chemical cleaning (also
called clean -in -place, or CIP) is necessary. As a general rule, chemical cleaning is
necessary when the system fouls to the point at which a 10-15% decrease in
normalized permeate flow, permeate quality, or pressure differential between the feed
and concentrate is observed.
A detailed discussion of the normalization of these parameters may be found in the
literature (AWWA, 2007; USEPA, 2005; USBR, 1998). Alternatively, many element
manufacturers can provide an automated spreadsheet tool for calculating these
normalized parameters based on routine operational inputs, such as flows, pressures,
temperature, and water quality data (e.g., conductivity or total dissolved solids). A
widely accepted general guideline for RO system operation is that cleaning should be
required no more frequently than once every three months.
It is important to note that most commercially available RO and NF products are
subject to damage by exposure to oxidants, such as most chemical disinfectants. RO
and NF membranes are generally not tolerant of free chlorine, which can quickly and
permanently degrade the membrane material, although some products may have a low
acceptable threshold for chloramine exposure. Consequently, these chemical
disinfectants are typically quenched or fully consumed in the treatment process stream
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 16

Guidelines for Cr(VI) Treatment Testing
prior to the membrane system. The element manufacturer should be consulted for
product -specific data on chemical compatibility.
Monitoring
Routine Monitoring
Water quality parameters to be measured in RO and NF pilot testing include at a
minimum: 1) Cr(VI) and total chromium; 2) general physical parameters, such as pH,
temperature, conductivity, total dissolved solids, turbidity, oxidation reduction potential
(ORP); and 3) SDI.
Additional water quality parameters should also be incorporated on a site -specific
basis, as -needed. Examples of such parameters are scaling ions that are identified as
potential foulants in the initial water quality modeling (e.g., calcium, iron, silica, barium,
etc.). Other water quality parameters may be added to the concentrate sampling
regime to address applicable limits relative to the method of management or disposal.
The primary sampling locations for high-pressure membrane pilot testing should
include raw water, pretreated water, permeate and concentrate (the latter two for all
stages if multi -stage RO or NF). Many of the parameters can be measured using on-
line instrumentation, such as conductivity, temperature, pH, turbidity, and ORP.
Frequent measurement (e.g., weekly for key parameters like Cr(VI) and total Cr to
develop a database of results) may be desired to characterize membrane rejection. If
frequent monitoring is not feasible, online conductivity measurements can be used as a
surrogate and trigger sampling if, for example, changes in conductivity rejection varies
by more than ± 25% over several hours. Less frequent sampling of the concentrate
should be conducted to characterize residuals and confirm the mass balance of
constituents removed by the membranes.
Operating conditions that should be measured include flow or flux, pressure, and
temperature, which are typically measured continuously.
Residuals Characterization
The membrane concentrate should be fully characterized for Cr(VI), total Cr, and any
other contaminants that may be regulated with respect to the disposal method
available. Regulated parameters may include not only toxics like Cr(VI), but even
general water quality characteristics such as turbidity and pH. The hazardous nature
of the concentrate should be assessed. If the concentrate is discharged to the sanitary
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 17

Guidelines for Cr(VI) Treatment Testing
sewer, specific water quality restrictions designed to prevent an upset of the
downstream wastewater treatment process might need to be considered.
Test Period
Membrane rejection of Cr(VI) should stabilize after a short period of operation, on the
order of days. However, it is generally recommended that the pilot system be operated
for a period of at least several months to optimize system performance, as well as to
assess the fouling rate. This period may be extended to accommodate variations in
water quality, if the source if subject to fluctuations over time.
III. Reporting
Reporting is a critical component to all Cr(VI) treatment technology testing to enable
appropriate use of the findings for subsequent planning and also for use as case
studies in future research efforts. The following components should be characterized
and reported for maximum usefulness of the information:
Testing objectives
o For example, providing an understanding of the goals of the testing,
describing unique features of the system considering treatment
including the desire to remove multiple contaminants, and identifying
reasons for testing innovative approaches
• System operations
o For example, influent water quality, system configuration, operating
conditions, results
• Unintended consequences observed (and those tested but not observed)
o For example, contaminant release or co -removal of other
contaminants by the process
• Residuals characterization and volumes generated
• Post -treatment considerations
o For example, pH adjustment for corrosion control
• Treatment technology integration in an existing treatment train
o For example, location compared with other unit processes and
additional needs associated with this integration
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 18

Guidelines for Cr(VI) Treatment Testing
A primary driver for thorough reporting is the desire to accumulate sufficient information
to inform cost estimate assumptions for treatment needs of a range of different utilities.
Cost estimates to date have largely been based upon one water quality and testing for
one utility, whereas those assumptions may not accurately reflect a good cross-section
of utilities' needs in Cr(VI) treatment.
IV. References
American Water Works Association. 1982. Design of Pilot -Plant Studies. Proc. AWWA Seminar.
American Water Works Association. 2007. Reverse Osmosis and Nanofiltration (2"d ed.) —
Manual of Practice M46.
Blute, N. and Kavounas, P. 2011. Cr(VI) Treatment Options. WaterRF Technology Transfer
Workshop on Hexavalent Chromium. 18 August.
Brandhuber, P. Frey, M., McGuire, M.J., Chao, P.F., Seidel, C., Amy, G., Yoon, J., McNeill, L.,
and Banerjee, K. 2004. Low Level Hexavalent Chromium Treatment Options: Bench -Scale
Evaluation. American Water Works Association Research Foundation.
Dow Tech Facts: Lab Guide. 2011. http://msdssearch.dow.com/PublishedLiteratureDOWCOM/
dh_003b/0901 b8038003b90e.pdf?filepath=liquidseps/pdfs/noreg/016-00003.pdf&fromPage=
GetDoc.
Kawamura, S. 2000. Integrated Design and Operations of Water Treatment Facilities. Wiley.
Lang, J.S., Giron, J.J.,Hansen, A. T., Trussell, R. R. and Hodges, W. E., 1993. Investigating Filter
Performance as a Function of the Ratio of Filter Size to Media Size. J. AWWA, 85(10), p.122-
130.
Malcolm Pirnie, 2008. Report on Additional RCF Pilot Testing to Optimize Design. Prepared for
the City of Glendale, California.
McGuire, M.J., Blute, N.K., Seidel, C., Qin, G., and Fong, L. 2006. Pilot Scale Studies of
Hexavalent Chromium Removal from Drinking Water. J. AWWA, 92(2), p.134-143.
McGuire, M.J., Blute, N.K., Qin, G., Kavounas, P., Froelich, D., and Fong, L. 2007. Hexavalent
Chromium Removal Using Anion Exchange and Reduction with Coagulation and Filtration.
American Water Works Association Research Foundation, Denver, CO.
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 19

Guidelines for Cr(VI) Treatment Testing
McLellan, N., McLeod, J., and Emelko, M. 2011. "'Wall Effects' During Filtration Investigations:
Reconsideration of Column -Diameter to Collector -Diameter Ratio Recommendations."
Proceedings of the AWWA Water Quality and Technology Conference. Phoenix AZ.
Qin, D., McGuire, M.J., Blute, N.K., Seidel, C., and Fong, L. 2005. Hexavalent Chromium
Removal by Reduction with Ferrous Sulfate, Coagulation, and Filtration: A Pilot -Scale Study.
Env. Sci. Technol., 39, p. 6321-6327.
Rad, S., Mirbagheri, S., and Mohammadi, T. 2009. Using Reverse Osmosis Membrane for
Chromium Removal from Aqueous Solution. World Academy of Science, Engineering, and
Technology. 57, p. 348-352.
Standard Methods for the Examination of Water and Wastewater. Published by the American
Public Health Association. 2005.
United States Bureau of Reclamation. 1998. The Desalting and Water Treatment Membrane
Manual: A Guide to Membranes for Municipal Water Treatment (2"d ed.). Water Treatment
Technology Program Report No. 29, July 1998.
United States Environmental Protection Agency. 2005. Membrane Filtration Guidance Manual.
Document No. EPA 815-R-06-009, November.
©2012 Water Research Foundation. ALL RIGHTS RESERVED. 20