Page 288 -

10.32 CHAPTER TEN

ways the case; because chlorine residual analyses drift off course unless they are regu-
larly maintained, feedforward control is better than feedback control alone.
Whenever chlorine demand is variable, compound-loop control is preferred. This sys-
tem is sometimes accomplished by controlling the differential vacuum regulator on the
chlorinator with the flow signal and setting the chlorine gas-metering orifice with the sig-
nal from the residual analyzer. Alternatively, it can be provided by electronically adding
both control signals and using the results to control the differential vacuum. The first ap-
proach allows the chlorinator to operate over a dynamic range of 200 to 1, whereas the
differential vacuum will only allow a dynamic range (maximum feed/minimum feed) of
20 to 1. However, in most installations, a dynamic range of 20 to 1 is satisfactory.
With either feedback or compound-loop control, lag time is one of the principal de-
sign parameters. Lag time is the time between the moment when the chlorine is added to
the effluent and the time when the residual analyzer signal comes to the chlorinator. Lag
time includes the transit time from the point where the chlorine is initially mixed until it
reaches the sampling point, the transmission time between the sample point and the chlo-
rine residual analyzer (in the sample line), and the analysis time.
The analysis time is usually a minor factor. If the lag time is too much longer than the
response time of the analyzer, the level of the chlorine dose will sawtooth. White (1999)
suggests that the lag time be maintained at an average level of 2 min, with a maximum
of 5 min. Low-flow conditions should be considered.
It should be understood from the beginning that the purpose of the chlorine residual
analyzer discussed here is to control the chlorine dose. If continuous monitoring of the
chlorine residual after the chlorine contact period is desired, another chlorine residual an-
alyzer is required.
The following are some of the most common design errors: poor chlorine sample con-
ditions, analyzer located too far from sampling point, and effluent chlorine dose paced to
influent flow.
Sample Point Location. The principal consideration in locating the sampling point
for the control analyzer is that there must be good mixing. If the sample is taken before
adequate mixing has occurred, the result will be erratic readings unsuitable for control.
For the majority of initial mixing designs, sampling should be provided immediately down-
stream of the initial mixing device. Ordinarily, chlorine residuals are stable enough for
control measurements after just a few seconds of contact.
If no initial mixing device is present, the sampling point should be far enough down-
stream to ensure that good mixing has occurred. For turbulent flow, 10 pipe diameters is
usually sufficient; however, low-flow conditions should be considered, and if adequate
mixing cannot occur in a reasonable time, an initial mixing device will be necessary for
control purposes.
Analyzer Location. Chlorine residual analyzers should always be located as near as
possible to the sampling point, even if special housings are required. Sample lines should
be designed for velocities of about 10 ft/s (3 m/s), and the transit time between the sam-
pling point and the residual analyzers should be minimized.
Chlorine Dose Pacing. The chlorine dose should always be paced to the flow most
representative of the point of addition. A common error in design is an arrangement in
which the effluent chlorine dose is paced using influent flow measurements. Too many
events occur between a plant's influent and its effluent, and such a design often results
in an erratic chlorine dose and an unmanageable operating system.

Chlorine Residual Analyzers
Two methods for continuous chlorine residual analysis are currently available: the auto-
matic amperometric titrator and the ion-selective probe. In an automatic amperometric