pH Control of Permeate Water

[polyroly]

Okay.....I am troubleshooting a plant that processes "dirty" water in a reverse osmosis plant and they have had issues regarding pH control of the permeate which is discharged to the environment.

Caustic is added before and after the water passes through the R.O. membranes to bring up the pH. Each line has a dedicated PD metering pump that doses in caustic based on the feedback from the pH probe on the permeate.

I have only toured the plant today and these are some things that I noticed that could be contributing to the instability of the pH control. Please give feedback!!

1. There is essentially the "old side" of the plant and the "new side" of the plant and they are mirror images of each other. The pH of the R.O. permeate is measured in a common sump which both sides of the plant connect to. I see this as a problem as this pH reading is supposed to control both sides of the plant.There is a static mixer after caustic is injected to each side of the plant and there is sample ports downstream of the static mixer but no pH probes?
2. The caustic is a 50% solution and they have had gelling issues. Tank is heated and all lines are heat traced. Could dilution help?
3. The pumps dose a small amount of caustic every 2 or 3 seconds. There is a long distance between the pumps and the injection points (upwards of 160 ft is the longest). There is no check valve at the injection point, is it possible that water is backflowing into the injection line? If so would this affect the caustic addition?
4. I am going to try and let the operator let me perform a step test tomorrow morning to see what the time delay of the pH control loop is.

Any comments or suggestions are welcome!! Sorry for the long winded background

Nathan

 

[shahyar]

2. When frozen crystalls apear, it is hard to revert them back in solution. Dilution does not help a lot.
3. There should be a "back pressure regulator" at the discharge side of caustic injection system. It is a pretty standard practice to make sure no back flow happens because of siphoning etc.
Hope this help.

 

[moltenmetal]

From what I can gather, you have far from an ideal set-up for pH control. I suspect that the permeate is weakly buffered so there will be a natural tendency to over-shoot pH at the best of times.

Is the sump mixed? If not, you've got a problem between the dosing and the response that you're measuring. If possible, consider piping the two lines together before they enter the dead volume of the unmixed sump and measure pH of the flowing stream.

Your pumps should have both stroke length and speed adjustment, with one being manual and the other (usually the speed or stroke rate) being modulated by the controller. If the pumps run infrequently and you're seeing cycles in the pH of your effluent, you need to back off on the stroke length or dilute your caustic- or replace with a smaller pump. The idea is to get the dosing to occur more continuously and hence more consistently.

If the pumps don't have both adjustments, junk them and get pumps that do. Good quality solenoid-actuated diaphragm pumps are likely cheaper these days than your time dealing with this.

A check valve with a high forward crack pressure (sort of a crude back-pressure regulator as noted by shahyar) will help ensure that caustic is delivered only when demanded by the controller.

 

[polyroly]

Thank you moltonmetal and shahyar

There is a backpressure regulator just after the discharge on each pump. The caustic that is injected into the permeate on the new side of the plant is injected into the bottom of the pipe (6 o'clock). This is the most critical location for caustic injection. Do you think a second backpressure regulator would be helpful in preventing siphoning?

 

[bimr]

1. Its seems a little odd that you are raising the pH prior to the RO system. It is more common to lower the pH of the feed into the RO prevent scaling of the membranes.

2. It is common to dilute the caustic prior to pH correction. One can feed the caustic neat or dilute, it is your preference. However, it is probably easier to control the pH using more dilute materials.

3. Feeding neat caustic may cause the temperature of the resulting solution to rise because of the heat of dilution. The resulting temperature depends on the flow rate of solutions.

4. If you are making a significant change in pH, you would be advised to accomplish in a small flash mixing tank rather than trying to accomplish this inline.

5. The pH of the permeate should be lower than the pH of the feed solution. That is because carbon dioxide is not rejected by the membrane and goes through the membrane into the permeate solution.

6. If you desire to control the permeate pH, you are advised that this is best accomplished after the RO system, not prior to the RO system. Adjusting the pH proper to the RO system should only be done to improve the efficiency of the RO process.

7. A backpressure regulator is commonly used to feed chemicals into process piping. You should also have a block and bleed valving system on the caustic line to prevent backflow of the caustic to the caustic tank.