393Arsenic Removal from Drinking Water by Ion Exchange and
Activated Alumina Plants
EPA 600-R-00-088
This report documents treatment plant information as well as results of year-long sampling
and analysis at two ion exchange (IX) plants (referred to as Plants A and B) and two
activated alumina (AA) plants (referred to as Plants C and D), with capacities varying from
800 to 3,000 gallons per day (gpd). The objective of sampling and analysis was to evaluate
the performance of the full-scale water treatment plants to consistently remove arsenic
from source water. Additionally, data were collected to evaluate the chemical
characteristics of residuals produced by these treatment processes.
The study was divided into three phases: source water sampling, preliminary sampling,
and long-term evaluation. Source water sampling was conducted to evaluate source water
characteristics at each plant. Preliminary sampling was initiated in August 1998 and
consisted of four sampling events conducted at each facility on either a weekly or biweekly
basis to refine procedures for subsequent events during the third phase. Long-term
evaluation consisted of weekly or biweekly sampling at each facility from September 1998
to September 1999.
Samples from resin regeneration were collected at Plant A from March to June 1999.
Spent AA samples were collected at Plants C and D during the media change-out events
in December 1998 and May 1999, respectively. Results from the long-term evaluation
demonstrated that both the IX and AA systems are capable of achieving arsenic levels of
less than 5 μg/L in the treated water, provided that the IX resin was regenerated or the AA
medium was changed out before arsenic breakthrough occurred. The two IX systems had
inlet arsenic concentrations between 45 and 65 μg/L [primarily As(V)]. When Plant A was
operated beyond 3,000 to 3,200 bed volumes (BV) of water, arsenic chromatographic
peaking occurred. Arsenic breakthrough was not observed at Plant B where an average
97% of removal efficiency was achieved, leaving only 0.8 to 4.5 μg/L arsenic in the finished
water.
Both AA systems consisted of two parallel treatment trains with a roughing AA column
followed by a polishing column in each train. The systems operated on a media throwaway
basis. The average arsenic removal efficiencies achieved at Plants C and D were 87%
and 98%, respectively.
The raw water at Plant C (34 to 76 μg/L total arsenic) contained approximately 0.3 to 28.8
μg/L As(III), which was nearly completely removed, even though no oxidation treatment
was provided. The water at Plant D contained slightly higher total arsenic concentrations
(53.3 to 87 μg/L) but no As(III), which was consistently removed to less than 5 μg/L in the
finished water.
The AA media in the roughing tanks were exhausted and disposed of about every 1 to 1.5
years after treating approximately 9,600 BV at Plant C and 5,260 BV at Plant D. The
regeneration process at Plant A recovered from 67 to 86% of arsenic from the spent brine.
The spent AA at Plants C and D passed the Toxicity Characteristic Leaching Procedure
(TCLP) test for metals including arsenic, and therefore was disposed of as nonhazardous
waste.