We are trying to neutralize demineralizer regen effluent. It is a continuous system that has varying flow rates and we do not have any flow indication. We are attempting to perform the control in an Allen-Bradley 5 PLC. I would like to develop characterizers for the SP and PV. The amount of neutralizing titrant needed to bring the process from a pH of 12 to 11, is 10,000 times the quantity needed to bring it from a pH of 8 to 7. We use 93% H2SO4 and 50% NaOH. Developing the characterizer is one problem. The tables I have are mL of titrant per volume of solution but without flow indication I don't know how much solution I have. Implementing the characterizer easily is another problem. Any suggestions?

 

The continous method of neutralizing the regen waste consume more chemicals than batch and expensive in long run.

If possible dump both base and acid waste into one tank and they neutralize each other to large extent and you save your chemical.
Another big issue to concern is the amount of resin that is carried away from the vessels to the treatment tanks. These act like a acid or base sponge and absorb a lot of chemicals before you can actually neutralize the waste. You see a pH 12 and you calculate and pump in acid and you see no change in pH. You do one more time and the pH will go to 2.

The retention time for chemical neutralizing is thus large for continous method of treatment.
More after listening from you on your views.

Best regards,
Sekar

 

Brett,
Wow, what a cool web site. Finally a social outlet for all of us "non-social" types, where we can discuss matters of true import.

I'm not an expert at pH control, so take my comments with a grain (or two) of salt. But I love discussing control problems.

Changing flow = changing velocity past the meter. This (and/or changing temperature) may effect the pH reading as much or more than actual changes in the pH, so beware, you may be chasing false upsets. Is there a collection tank downstream? 2 minutes resident time would be GREAT! So, if there is, I recommend moving the pH probe to the tank. Or better yet install a second probe in the tank and use the line reading to generate a "feed-forward" bias and control on the tank pH.

Note that your acid and base provide markedly different equivalent concentrations, which would prompt me to use separate loops on either side of 7.

Tell me more about the process and let me know if you need help setting up the loops in the PLC-5.

Take care,
Jeremy

 

Your neutralization is similar to water treatment problem. The plant utilizes Residual Only control to complete the process. A fixed lag time can be associated with this control and should = the dead time of the control loop associated with updates to the pH (PV). Use either a Time.Delay.On block to force PID updates or take the lag time into account when developing the reset time (Integral gain).

Another approach includes an Integral Gain schedule, which changes your reset time based on deviation from set point (or error). When the error is within 1 pH, use a longer reset time whereas when the error is 5 pH, use a shorter reset time.

 

A couple of suggestions.

1. DI waste is always best neutralized in a batch mode. Send the waste into a tank, wait until its done regenerating, then neutralize it all at once.

2. You can usually get close to net neutral pH by adjusting the volume of acid or caustic used to regenerate with. It matters not whether you
introduce the chemicals as part of the regeneration sequence or into the waste tank itself. If nothing else introducing it via the regen sequence at least runs it through the resin bed potentially improving the quality of the
regen, where as putting it into the waste tank does nothing for the regen itself. It may tank a few regens to adjust the regen chemical dosage to get close to a neutral pH, but its the best solution. Much easier to get from 9 to 8 than from 10 to 8.

3. if you have substantial amounts of resin in your waste stream you have a serious problem with your equipment that needs to be repaired. You will get a few fines out during backwashing (if your system needs that step in the regen sequence) but if you are getting resin in the waste stream in any volume, you need to fix that problem first.

4. Consider recycling some of your waste water so you don't have to neutralize so much. The backwash water can often be sent back to a
pretreatment stage (such as a clarifier or gravity filter). Some of the rinse water may also be reusable. Adding a diversion valve to the waste stream and a conductivity monitor to control the diversion valve is pretty cheap and can save you having to neutralize lots of water. Plus you save on not having to make the water in the first place. This assumes you have some
place to send the recycle water to (such as a pretreatmetn system). You could also add a recycle pump for your final rinse step rather than sending it to drain.

Bob Peterson