Common errors with pH sensors
pH Things that go bump in the night Story time! Once upon a time, I had to troubleshoot a pH measurement in effluent water from a wastewater plant.
pH Things that go bump in the night
Story time! Once upon a time, I had to troubleshoot a pH measurement in effluent water from a wastewater plant. I knew the technicians had properly installed the loop, consisting of pH electrode, cable, and transmitter. A top-down mounted holder fixed the electrode to the side of the outlet channel. Only the sensor came in contact with the water; the cable and contacts stayed safe inside the holder. But they asked me to find out why the plant manager called the measurements “time dependent.”
At night, they found discrepancies between the values from the measuring point and the sample analysis done in the lab. However, during the day, the values matched properly. We excluded human error because the same staff did the analyses. Installation and equipment also checked out, so we changed the pH sensor.
Things worked well for a couple of days, but then the problem recurred. So I visited again to dig deeper. During the day, the values corresponded in the range of 8.4 to 8.5. At night, the online value dropped to 7.1 while the lab value stayed at 8.4. Why the drop?
You may have guessed by now that it had nothing to do with the analysis. It was the contact resistance between the cable and the pH sensor. At pH 7, the millivolt (mV) signal is close to 0. Then for each step in pH, the signal changes 59.5 mV. This means that an electrode in a neutral solution at pH 7 gives close to 0 mV, pH 8 around -59.5, pH 9 gets -119, and so on.
During the day, the cable and the sensor had good contact and sent the complete signal to the transmitter. At night, the humidity and temperature changed enough to create resistance on the contact and reduce the signal.
Cable and contact issues often create hard-to-detect problems with pH measurement. You have to know what to look for – or eliminate the possibility altogether.
Digital pH electrodes will never cause the above issue. They work like regular electrodes, with measuring and reference half-cells. But digital electrodes have small microprocessors in them that convert mV signals into digital signals that shrug off environmental interference like contact resistance.
I could go on and on about possible issues, but you don’t need to worry about all of them. In my experience, pH measurement is simple to perform with reliable equipment. Still, I’d like to bring up another common issue and how to get around it. Maybe you’ll see a clue and solve the case before I finish!
Mr. Technician in the plant with the sample
In this mystery, a pulp plant had serious discrepancies between the pH values from the process equipment and the lab. The values matched almost randomly, but often the process measurement fell about one pH unit short. Recalling what I learned at the wastewater plant, I examined all the connections, but they checked out okay.
I also considered galvanic isolation, hard to detect and common with conventional pH loops. Basically, different electrical potentials may occur between the grounding of the device, the control system, and the pipes where the measurements take place. Any extra electrical potential can interfere with the electrode’s tiny signal.
An easy way to detect this issue is to take a grab sample and measure it with the same device used for the process. If the values differ, then you should look for a ground loop in your plant. In this case, that wasn’t the issue, even if I got different values between the online and offline measurements.
I discovered it had to do with the sample from the pipe. If I kept the sample’s flow rate low from the valve through the sample tube into my beaker, I got values that matched just fine. If I flushed the sample into my beaker, the sample changed, probably because of evaporating carbon dioxide. Suggesting a smooth grab-sampling procedure closed this case neatly!
Keeping cool when temps change
We all know temperature can affect measurements, but many users think that if their pH meters have automatic temperature compensation (ATC), then they can ignore their sample temperatures. Wrong.
You can’t compensate for the temperature dependence of your sample unless you do a lot of programming on your pH transmitter. Each sample will have its own dependence. Remember the equilibrium formula for an acid? Time to roll the clock back to our school days!
The more reaction you get from the right, the more H+ ions you’ll have in your solution and thus a lower pH. How much of the reaction goes right or left depends on the equilibrium constant Ka, which is temperature dependent. So a sample can have different pH values, depending on its temperature.
Now you also understand why it’s tough to compare pH values in a process running at 80 degrees Celsius and a grab sample in the lab at 25. Both values are right, just measured at different temperatures. So always note the sample temperature when you measure a pH value.
The two faces of ATCs
Now ATCs have two purposes:
1. Compensation for electrode temperature dependence – The half-cells each have a dependence and will give you errors if you don’t adjust the sensor. The slope value, or sensitivity to pH, is around 59.5 mV per pH unit at 25 degrees Celsius for a good sensor. At 80 degrees Celsius, you get a value of 70.1 mV. A process running at 80 with a pH of 9 would mean an error of 0.8 pH.
2. Easier sensor adjustment – Look at the pH-buffer bottle you use for the sensor calibration. You’ll see a temperature and pH table printed on it because each solution has its own temperature dependence. You need to know the buffer’s pH value by measuring the temperature, especially in alkali buffers with pH 9 or 10. If you use an ATC programmed to the correct buffer, the sensor will know what value to use. Just remember to program your device correctly!
As I mentioned before, you can run into all kinds of other problems with pH sensors, but they’re pretty rare. These are the most likely issues you may face, and now you know how to tackle them!