Chlorine and Sod Bis control


Thread Starter


I am trying to find a control philosophy for the control of chlorine dosing pumps with variable speed and stroke control.

I have been told that PID control is not suitable.

Also Sodium Bisulphite control is required which can also be tricky.......

Any comments appreciated


Darran Weissenborn

Hi Pete,

I have used many different systems in the past to achieve repeatable dosing of chlorine, PAA, etc.

The trick is to ensure that the dosing rate always stays proportional to the flow rate of the medium being dosed!

The most accurate method I have used to date is by installing a flow meter down stream of the injection point and use its pulsed output to drive your dosing pump.

I was able to control to +/- 20ppm available chlorine using this method.

Hope this helps!


Darran Weissenborn

J.J. Overyton

If you have the answer let me know also, we have the same problem at a waste water plant, does anyone know of a good sod bis analyzer.

Blunier, Mark

What are you using for an online sensor? We add Sodium Bisulphite to neutralize chlorine, but it is metered in at a set rate, and adjusted per lab tests.


Any opinions expressed in this message are not necessarily those of the

Al Pawlowski

PI control is suitable and works well when you combine it with flow feed forward. I like to use a ratio controller for the flow with the ratio set by a PI controller for the chlorine residual. Bisulphite can be ratioed off of the chlorine feed (output of the chlorine cascade if you don't have a chlorinator rate valve position output). BTW, this setup is commonly called "compound" control.

This scheme works well if the physical part of the process is set up right. Usually, the tricky part of the whole is getting good injection and monitoring points so that you don't have large lag times between residual measurement and feed (greater than about 10-15 minutes) and the lag does not change too much with varying flow.

I have always wanted to try a Fuzzy control scheme on this application. Seems like it might offer a real improvement.

Darold Woodward

I have achieved this successfully in several projects. Here are some details.
Please be warned that this isn't a "design" and your exact plant conditions, permit, and other factors must be carefully considered before proceeding.

PID is unsuitable because of windup and delay in Cl2 residual readings. However, you can use a combination control to acheive a pretty good outcome.

In one plant, the operations staff did not believe that automated control was possible so we provided a "pacing factor" control. We also were required to meet a zero Cl2 effluent requirement. The Na(SO4)2 was dosed on a flow
proportional basis. Operators then observed trends of Cl2 residuals in the contact chamber and effluent reduction potential. If Cl2 was satisfactory, small adjustments were made in the pacing factor to always maintain an effluent reading that indicated that there was some remaining bisulfite.

Similarly, we designed a fully automated system using a compound control. A simple flow-pacing is used for bisulfite with a "trim" factor of + or -. The trim was limited to a small range (maybe 10%) and allowed to bias the output of
the flow pacing. The trim was based on a PID control using ORP. Windup can still be a problem, so careful tuning and testing are important.

You could also use Cl2 residual to drive the proportional part, but that requires proper placement of the Cl2 instrument and use of the proper instrument. It is important in wastewater to use a total Cl2 residual instrument rather than free Cl2 because of the chlorine uptake resulting from ammonia present in wastewater streams.

One of the mfrs in the wastewater business swears by ORP (Oxydation Reduction Potential) for control of both Cl2 dosing and dechlorination. They make package systems for chlorination/dechlorination at small plants. ORP can be useful since it is supposedly more closely related to kill than Cl2 residual and doesn't require a reagent in the instrument (expensive and difficult to maintain). An ORP sensor may not be sufficient for your reporting unless you
are only reporting grab samples. ORP is used in a compount control similar to what I described above.

One of my colleagues from those jobs is on the list also and worked with several systems... any comments Don?

Good luck...

Darold Woodward PE
SEL Inc.
[email protected]

Robert Dusza


Try using ORP for both. We have been using if for about 6 years now with few problems. We have gas CL2 and SO2 (dechlorination) but the principles are basically the same. We use PID for the control, just need to know the lag time and use a baseline feed of the chemical to ensure you donot drop to zero outout of the product.

The treatment plant in the next town is using liquid for both processes and using PID for control. I have some info on controlling CL2/DeCl2 systems that may be helpful.

Hope this helps.


Robert J. Dusza, Jr.
Treatment Manager
(V) 1-860-647-3219
(F) 1-860-647-3150
E-mail - [email protected]
Manchester Water & Sewer Dept.
125 Spring St. P.O. Box 191
Manchester, CT 06045-0191