Chlorine Pacing for Water Treatment Plants

D

Thread Starter

Don Best

What kind of control strategy is most often used to pace chlorine (Sodium Hypochlorite) at a Water Treatment Facility or Reservoir? In my limited experience, I've only seen flow-paced or open-loop control (with a "pacing factor") used for this. Could someone use a Residual Chlorine Analyzer with a PID loop controller to pace the speed of a metering pump to maintain a desired Chlorine Residual setpoint?
 
We did Sodium Hypo water treatment for Atomic Energy of Canada, in 1980. The only gear capable was Micon. The chemestry is not obvious.
The analyzer was Orion.
 
A

Alan Rimmington

I have a number of sites [more in pipeline ;-)] working using residual feedback - combined with a correction for changes in flow. I would advise not using a "standard" PID control block otherwise problems could occur with control stability.

[email protected]
 
T

Tom DeFreitas

Most of the systems I program will have two modes of control and the operator can choose between them. 1. Flow paced 2. Flow Pace with residual trim.
I always use the proecss flow rate as the feed forward/bias for the residual trim mode. Many times there is a significant deadtime associated with process so your PID loop update time must account for it. Using the FF/Bias keeps the PID loop from having to adjust based on the flow changes and the residual trim controls the chlorine demand of the process.
-Tom
 
J
Don Best wrote:
>What kind of control strategy is most often used to pace chlorine (Sodium
>Hypochlorite) at a Water Treatment Facility or Reservoir?

Chlorine pacing is usually done by a chlorinator. This is a combination of a feed pump, analyzer, and PID controller. I suppose it is done this way because of the importance in maintaining the right CL2 residual - there are risks attached to placing the chlorine injection system into the general purpose control devices that will likely be buggy during the plant startup and testing.

There is nothing stopping you from rolling your own though. The control algorithm to use is flow pacing with PID trim.

Many chem feed pumps have an independently adjustable stroke and speed. Make the speed proportional to the plant flow and regulate the stroke with a PID loop where the residual CL2 is the process variable and the desired dosage is
the bias. The loop can be tricky to tune because of the iteraction between the plant flow rate and the time it takes for water to move from the injection point to the measurement point.


Jay Kirsch
[email protected] <mailto:[email protected]>
 
D

David Holman

Don,

A method I found that worked in a waste-water treatment plant for chlorine
dosing was to combine the two methods.

A PID loop was used to control the dosing rate (PV = chlorine residual, SP =
desired residual, CV = dose rate). This loop was very slow due to the long
detention time in that application.

Flow pacing was used to control the speed of the dosing pump:

Dose (pump speed) = Dose rate (determined by PID loop) * flow rate * Pump
Calibration.

David Holman.

Holman Technologies Pty Ltd
 
W
Most chlorine feed applications use only proportional control, not PID. Having the ability to do PID control is usually a blessing, but very few water or wastewater treatment plant operators have that level of instrumentation knowledge. Just look at the incident in Ontario CA last year.

Chlorine gas is much easier to control, but there is so much FUD about it that hypochlorite, and even onsite generation are becoming more popular.

Hypochlorite feed is difficult to control via feed forward methods, either
flow-paced or residual paced. Generally the best type of control is a
compound loop, with the main proportionality being flow paced, and the trim
being controlled by residual.

Sometimes it is useful to stage multiple metering pumps at a constant feed rate, especially if the flow rate is changing in steps, such as pumping out of a wet well.

Walt Boyes

---------------------------------------------
Walt Boyes -- MarketingPractice Consultants
[email protected]
21118 SE 278th Place - Maple Valley, WA 98038
253-709-5046 cell 425-432-8262 home office
fax:801-749-7142 ICQ: 59435534

"Strategic marketing, sales and electronic
business consulting for the small and medium-sized
enterprise: http://www.waltboyes.com"
---------------------------------------------
 
W
The chemistry _is_ obvious. Just go to the WEF or AWWA and get a good text on chlorination.

Doing this from memory, now...

Simply, hypochlorite dissociates in water to make hypochlorous acid and hypochlorite ion. Residual analyzers measure either "free chlorine" which is the amount of hypochlorous acid and hypochlorite ion in the water, or they measure "total chlorine" which includes the sum of all free chlorine plus "combined chlorine" (chlorine and ammonia compounds, THMs etc.)residuals.

Most water treatment plants are interested in "free chlorine residual" since it is the most powerful disinfectant.

Most wastewater treatment plants are interested in "total chlorine residual"
since they usually have plenty of ammonia, and combined residual lasts a very long time.

Sometimes water treatment plants deliberately feed ammonia to produce chloramines for specific purposes.

Residual is defined as "dosage less demand".

Controlling chlorine dosage is the practice of dosing a small enough more
chlorine than is necessary to meet demand.

The problem is that chlorine is not very soluble in water, and tends to come out of solution easily, and mixes poorly.

Another problem is that it is not possible to directly measure "demand." You measure demand by measuring residual after an initial contact period, which is usually 30 minutes. If you use this 30 minute physical loop lag to modify your control system, you are going to have lots of fun.

So, not only do you have to have a decent control algorithm, but the physical placement of the injection point and the sample point are critical to performance. Do it wrong and you will get disease through the plant.

Back in the early 1980's, the EPA released a report that said that the automatic controls were turned off on over 80% of chlorination systems reporting to them. My belief is that it isn't much better now.

Walt Boyes
 
Thank you for this post. It helps me a lot on how to control the pump speed and stroke for sodium hypochorite pumps.

I only have clarification on what is that "Pump calibration"? Can you give me example of pump calibration value with engineering unit?

I just want to see how to calculate the speed using the formula. I have the following range with engineering units:

speed = 0-100%
Stroke = 0-10mg/L
flow = 0-300 gph

Thank you in advance.

> A method I found that worked in a waste-water treatment plant for chlorine dosing was to combine the two methods.

> A PID loop was used to control the dosing rate (PV = chlorine residual, SP = desired residual, CV = dose rate).
> This loop was very slow due to the long detention time in that application.

> Flow pacing was used to control the speed of the dosing pump:

> Dose (pump speed) = Dose rate (determined by PID loop) * flow rate * Pump Calibration.

> David Holman.
> Holman Technologies Pty Ltd
 
B
The thing is that you need to decide what your dosage rate per GPH is and work backwards from there.

I don't understand what a stroke of 0-10 mg/L means. Normally the stroke is so many liters or milligrams or whatever unit of volume is used per hour when the pump stroke is at 100%.

Usually, you are trying to does to some level of free Chlorine. Running open loop like this won't always get you there depending on the water quality.

--
Bob
http://ilbob.blogspot.com/
 
Some great replies here. Obviously experienced hands.

I would only emphasize a couple of points made earlier that are very important:

1. remember that the chlorine feed may not show up in the analyzer reading correctly unless the analyzer is far enough down the pipeline to allow for sufficient contact time/mixing. and if it's far enough downstream in the pipeline, the effect of a change in feed rate will not show up quickly, so your loop should account for this delay.

2. you should definitely roll your own feedback loop as a non-standard loop. and try to add setpoints for min feed rate and max feed rate, so that your chem feed loop doesn't do anything too crazy like an unrealistic overfeed or underfeed rate.

-Jim Mimlitz, SCADAmetrics
 
Sorry Bob, correction:

Speed = 0-100%
Stroke = 0-100%
Total residual Chlorine = 0-10mg/L
Flow = 0-300gph

Yna

> The thing is that you need to decide what your dosage rate per GPH is and work backwards from there.

> I don't understand what a stroke of 0-10 mg/L means. Normally the stroke is so many liters or milligrams or whatever
> unit of volume is used per hour when the pump stroke is at 100%.

> Usually, you are trying to does to some level of free Chlorine. Running open loop like this won't always get you
> there depending on the water quality.
 
Most of the systems I program will have two modes of control and the operator can choose between them. 1. Flow paced 2. Flow Pace with residual trim.
I always use the proecss flow rate as the feed forward/bias for the residual trim mode. Many times there is a significant deadtime associated with process so your PID loop update time must account for it. Using the FF/Bias keeps the PID loop from having to adjust based on the flow changes and the residual trim controls the chlorine demand of the process.
-Tom
Hello, Tom.
Do you happen to have a screenshot of the logic you have developed. Greatly appreciated.
 
Don,

A method I found that worked in a waste-water treatment plant for chlorine
dosing was to combine the two methods.

A PID loop was used to control the dosing rate (PV = chlorine residual, SP =
desired residual, CV = dose rate). This loop was very slow due to the long
detention time in that application.

Flow pacing was used to control the speed of the dosing pump:

Dose (pump speed) = Dose rate (determined by PID loop) * flow rate * Pump
Calibration.

David Holman.

Holman Technologies Pty Ltd
Hello David,
What do you mean by pump calibration, would that be pump capacity?
Thank you for your help.
 
Top