Cooling water pH control


Thread Starter

Manoj Topiwala

Hi List !

Presently we are controlling the pH of cooling water by dosing H2SO4 manually based on the indication of on-line pH meter in the supply line of cooling water.

Now we want to automate the pH control with help of local pH indicator/controller (or through DCS) & auto on/off of dosing pumps.

We would like to know if anybody had done this kind automation for cooling water pH control. What are the precautions to be taken & problems faced.

Thanks in advance.

Manoj Topiwala

Jerry Robertson - BTS Systems, Inc.

Well, we have not done it on cooling towers but we have done in parts washers. Similar as we are trying (successfully) to control pH and conductivity. Our approach was to use a simple PLC from Sixnet and program a really simple control loop that looked at pH and allowed a small deadband before we added chemical. We also have the option of controlling chemical based on time instead of pH. Handy for quick changes. We make a custom user interface to control and log data at user defined intervals. This system has been installed on several parts washers and had good success controlling the parameters we were trying to control.

A couple of things, the sensor for measuring pH can not be right on top of where you dump in the chemicals. What happens is that the sensor sees a big change and shuts off the controls dumping in chemical too soon. You also have to allow for the lag time if you are located too far away. A little tweeking will get it.

If you are trying to measure chemical usage you will have to carefully calibrate your pumps too. One thing you will find is that your chemical usage will go down and the pH level will stay more constant if you get the loop control right.

Feel free to contact me off-line as I don't hang around the list too often. I am going to be traveling for the next couple of weeks but I should be able to get to my emails. Out of the country but I think I have an ISP all lined up. I hope. My email is [email protected]

Jerry Robertson

Mihir Ramkrishna

Hi Manoj
I guess, the loop shall work very much in the same way as the normal control loop because most of the pH transmitters will give 4-20 mA compatible with DCS or controller.

I have a suggestion to make. I suggest u to use two pH sensors in place of one. You can develop suitable software on the DCS side so that it will consider the good values only. I think if it is critical loop then this should help u.

Bob Peterson

A common issue. A few suggestions.

1. Make sure the range of your pump is adequate across the full flow and pH range of your feed water. Remember that the pump will have to turn down by a factor of 1000:1 to control across 3 decades of pH (e.g.-10 to 7 pH). Varying feedrates can make this even worse. you may need more than one pump to make this work with one pump being used for low injection rates, and
another for higher injection rates.

2. Try not to be perfect. This is not typically a critical application, so if it is close, leave it alone.

3. Feedforward based on both your feed rate and the incoming pH works a lot better than trying to control with just feedback.

4. Try to linearize your pH feedback signal. keep in mind that a pH of 9 is 10 times more basic than a pH of 8 (with respect to a pH of 7). Thus you should probably use an inverse log function on your pH feedback and control off of that instead of off the pH directly. a pH of 7 really means 10^-7, and a pH of 8 is really 10^-8, etc.

5. Once you get closer to neutral pH you will find that very small changes in output produce very large changes in pH. A setpoint of 6 or 8 is usually a lot more stable than one at 6.5 or 7.5.

6. pH controllers are available that already implement these ideas, so unless you have some reason you feel the need to roll your own, its best to buy the solution.

7. Sometimes its necessary to titrate the water being treated to find out just how much acid you will need to add to get the right pH, and this can be added into the algorithm. However, this is a good solution only when the water being treated does not change much and typically cooling water is not in this category.

Bob Peterson

Ian Verhappen

I would suggest that you purchase Greg McMillan's book on pH control from ISA ( as a starting point as he described many do's and do not's for pH control. Two key things I would suggest are:

1. Since the pH reading is logarithmic, be sure to have accurate dosage metering and Positive Displacement dosing pumps.

2. Do NOT use 2 pH meters. If you are planning to install multiple meters go with 3 since if you use 2 how will you know which one is correct when they don't agree most of the time.

Ian Verhappen
[email protected]
Dear Manoi,

It's been a while, but I believe we implemented an adaptive gain strategy and dead time compensation for pH loops. Usually there is a signal delay from where the acid is injected to where the measurements are taken place, and your controls must account for this in the form of a Smith Predictor or like algorithms.

Also the response of the pH variable can be very non-linear and touchy around the setpoint so we adjusted the gain in this region to suit.

Feed-forward strategies may also be of use depending on the situation.


Paul Jager, President AutomationX Industrial Solutions (250) 724-1402
Somebody already said it. This is not rocket science. There are likely 20 or more companies that provide pH control systems and many of them are specifically in the cooling tower business. Walchem, Liquid Metronics, Pulsafeeder, Danaher/GLI, GEBetzOzmonics, USFilterStranco, and many many more...including Emerson, Foxboro, etc.

Contact a vendor, or a water treatment company like GEBETZwhatevertheirnewnameisgoingtobe, or Nalco, or even your local water doctor, and get them to specify the equipment and go do something with more value added for your engineering time.

Walt Boyes

---------SPITZER AND BOYES, LLC-------------
"Consulting from the engineer
to the distribution channel"
[email protected]
21118 SE 278th Place
Maple Valley, WA 98038
253-709-5046 cell 425-432-8262 home office