pH probe for treated water

[OilBoiler]

Hi everyone,

I am designing a caustic injection system to our wastewater system. Currently, the line is a 24'' line with a flow of about 2500-3500 gpm. This is water that has already been treated goes to a tank and then to the river. There is an inline mixer in place already. The idea is to inject caustic upstream of the inline mixer, have a feedback pH loop downstream of the mixer: with a pH analyzer along with a controller that will send signals to the caustic control valve and adjust caustic addition. I have two main questions:

1. Is the feedback loop better for this service rather than a feedforward controller? The pH analyzer for the feedforward mode would be placed upstream of the caustic injection and would send signals to the control valve.

2. For the pH analyzer: this is a 24'' line with relatively low velocity and pressure (1-2#). Would the analyzer read a more accurate pH if it is placed in a 1'' bypass loop to this line? Or is it feasible to install the probe directly to the line? (The budget for this project is quite tight). I guess my question is what would be the criticality of having an accurate reading whether it is installed at the 24'' line or at a 1'' bypass line.

Thanks in advance!

 

[ICman]

1. A feedback loop is simplier and should work just fine. It may depend on how well the in-line mixer works, though, but that can probably be overcome thru tuning.

2. A bypass loop will be slower and more expensive and possibly less accurate than a direct measurement. You can purchase probes that can be mounted directly on the line fairly inexpensively. However, you cannot use these in certain circumstances (high pressure, large particulates, slurries, anything that could foul or break the probe, etc.). Though your pressures are low, I would still install valve(s) so you can remove the probe(s) without making a mess.

Hope this helps,

. . . Steve

 

[m777182]

I disagree with ICman's answer in point 2: a bypass line for the probe allows you to maintain or replace the probe more easily. You can splash it from time to time, you can insert some bachwash filter and you can calibrate or check the loop automaticaly by injecting some buffer into the bypass line.
m777182

 

[ICman]

I agree that the bypass line allows easier maintenance (usually), but OilBoiler said the budget was "quite tight". Unless the piping configuration allows you to install a bypass line (say at an elbow), it could be expensive to do.

You may still have to install valves to isolate the line. You may not get a true representative sample, and you can pick up some sediment that could foul the probes. The temperature could be slightly off, but probably not enough to affect the measurement.

I have found that it is pretty easy to pull a direct-insertion probe from a line to check it, clean it, or stick it in a buffer solution for calibration.

Disclaimer: Most of the applications I worked on were chemical processes, not wastewater. Bypass lines had to be rerouted back to the main line or to a separate waste line, and could not be simply put down a drain. [flush]

You have a good point, m777182. In this case, however, a plastic 1" bypass line to a drain may be a viable and cost-effective solution. OilBoiler will have to decide what he can do within his budget.

 

[CARF]

OilB,

pH-control can be nasty thing since it's highly non-linear.
Of course feedback control is the best option. Feedforward on pH or flow can be applied to improve response of the feedback control loop.

What we did is have two parallel lines, with a pH probe in each line. Both can be calibrated and checked independently. If the values of the probes differ to much an alarm is generated.

Be very keen to specify the required control performance:
i.e pH-setpoint +/- 1 pH value

Try to log data on this system to analyse the preformance of the loop.

http://www.pfaudler.com/gle/instmnt.htm

Good luck and try to raise some serious budget,
CARF