In November 2000 the Washington Aqueduct, which is the water treatment
provider to the District of Columbia, switched its disinfectant from chlorine to
chloramines in order to reduce disinfection byproducts (DBP). However, accompanying this change was an increase in 90th
percentile lead levels from 12 parts per billion (ppb) to 70 ppb. Lead profiles conducted by District of Columbia Water & Sewer Authority (DCWASA) showed average peak concentrations up to
150 ppb in homes with lead service lines. (Giani & Edwards 2004). Review of the
profiles indicated that the majority of the lead was dissolved and coming from the lead
service lines.
During the spring, the Washington Aqueduct conducts an annual chlorine burn
for five consecutive weeks. During these five weeks, chlorine is added as the disinfectant
with free chlorine residuals maintained at approximately 3.5 mg/L in the distribution
system. In 2004, the chlorine burn initiated in April. During this time, lead profiles
conducted at customer homes showed a dramatic decrease in peak lead levels, < 15 ppb
in a few weeks time frame, (Schock & Giani?2004). After the "burn" was
completed, chloramine was reinstated. Throughout the next few months, peak lead
profile concentrations began to increase well above the action level again. During this same time frame, lead service lines were removed from the ground
and shipped to the US Environmental Protection Agency's research laboratory in
Cincinnati, Ohio. X-Ray diffractions revealed that the majority of the scale consisted of
lead oxide (Pb -IV), (Schock & Giani - 2004). Schock also revealed that the Pb-IV
could have formed due to high oxidation-reduction potential (ORP) caused by using
chlorine concentrations prior to 2000 in excess of 3.5 mg/L (free chlorine) in
the DCWASA system. When chloramines were introduced, the ORP decreased thus
causing the lead oxide to dissolve. Beginning in March 2003, DCWASA engaged in a series of pipeloop studies to
determine the most effective optimal corrosion control treatment. During these studies, it
was determined that the addition of orthophosphate was the most optimal treatment. It
was also discussed that potential switch back to chlorine might help speed up the lead
reduction process.
Several pipeloops were being utilized at DCWASA's Fort Reno Water Quality
facility to assist in optimizing corrosion treatment. Two of these pipeloops (Pipeloops #3
and #6) were dedicated to study the effects of chloramines and phosphate, chlorine and
phosphate and the switching back and forth of the disinfectants and their effects on lead
release from the lead service lines. It was also a concern to study the effects a chlorine
burn might have with regards to lead release once the phosphate has had a proper chance
to passivate the lead.
This paper discusses the findings of pipeloops #3 and #6 with regards to lead
leaching concentrations under the scenarios previously mentioned. Includes 3 references, figures.