Page 271 -

OXIDATION AND DISINFECTION 10.17

ramine residual permitted at the far end of a distribution system depends upon individual
state regulations. However, a residual of less than 0.5 to 1.0 mg/L will make the system
vulnerable to the onset of biological nitrification. Once nitrification starts to occur, the
only way to correct this problem is to change to the use of a free-chlorine residual until
the nitrifying microorganisms have been inactivated. Some utilities switch from a com-
bined to a free-chlorine residual at regularly scheduled intervals to prevent the onset of
nitrification. The placement of rechlorination stations at strategic locations in the distri-
bution system can also be used to reform monochloramine from the ammonia released
during the decay process.
Adding refrigeration-grade anhydrous ammonia, ammonium sulfate, or ammonium hy-
droxide to water containing free chlorine will form chloramine. Both the chlorine and the
ammonia must be very rapidly mixed into the complete volume of water to prevent the
formation of trace amounts of di- or trichloramine. The chemical feed system must be de-
signed to maintain a constant ratio between the free-chlorine residual and the quantity of
ammonia-nitrogen being added to the system.
A comprehensive water quality monitoring program should be conducted throughout
distribution systems using a chloramine residual. Parameters measured should include
combined chlorine residual, free ammonia, nitrite and nitrate ion, as well as the bacteri-
ological and physical measurements routinely performed on these samples. The most sen-
sitive indicator of the onset of biological nitrification will be the presence of nitrite ion
in the sample.
Monochloramine provides a long-lasting residual in distribution systems, but it is in-
herently unstable even in the absence of reactive substances. The net reaction, simplified
from the approximately 14 individual reactions that govern it, can be written as

3NH2C1 = N2 + NH3 + 3C1- + 3H +
A kinetic model describing chloramine formation and auto-decomposition has been
developed by Jafvert and Valentine (1993) and Vikesland et al. (1998). Use of this model
is relatively complicated. However, in the absence of chlorine demand reactions, mono-
chloramine decay can be estimated by a simple second-order equation. The integrated
form of this relationship is

1 1 _ kVCSCt
[NH2C1] [NH2CI]o
where ~v'CSC is the Valentine chloramine stability coefficient. This coefficient increases
with decreasing pH and the initial chloramine concentration that determines the total free
ammonia present in the system. It also increases with increasing total inorganic carbon
and temperature of the water. Utilities may use this coefficient to calculate the effects of
water quality and chloramination practices (say, mg C12/mg HN3) on monochloramine
stability. Observed decay rates that exceed those predicted by the stability coefficient can
be used to locate problems in the distribution system, such as the presence of oxidizable
iron or organic slime on pipe walls. Observed decay rates that are less than the stability
coefficient may point to the existence of relatively stable organic chloramines in the water.
The chemistry of chloramination in the presence of bromide is not completely under-
stood. However, it is well documented that several compounds may be produced by the
reactions between chlorine, ammonia, and bromide. Both bromamines (NHzBr, NHBr2,
and NBr3) and bromochloramine (NHBrC1) have been found (Wajon and Morris, 1980;
Symons et al., 1998). The Metropolitan Water District of Southern California reported an
increase in TTHMs from 10 to 20 ~g/L while using chloramines during a drought when
the bromide level in the raw water increased to 0.5 mg/L.