Like the old saying goes: "Give a man a watch and he always knows what time it is. Give him two watches and he never knows the correct time." However, give him three watches, let him average the two closest times, and he'll be fairly close to the right time. The same can be said of pH. Using three sensors leads to more accurate, dependable and believable measurements.
Let's start with the basics of how to measure pH accurately. The first thing needed is an accurate reference or starting point — that's the pH buffer. This is the standard by which a pH sensor is calibrated. The sensor is cleaned (hopefully) and then immersed into buffer solutions of 7 and 4 pH to set its zero and slope. Properly calibrated this way, the sensor will be able to measure correctly.
But what happens when the process does what it does so well — coating or otherwise adding error to the sensor measurement? What about potential ground loops that may or may not arise on any given day? How do you ensure the pH value you are seeing is correct?
Some plants rely on a grab sample (Figure 1), running it to the lab for a quick comparison. However, getting the sample to the lab takes time, which can alter the temperature of the sample or allow it to absorb carbon dioxide from the atmosphere. Both of these will make a lab reading differ from that of the online sensor.
You also must consider a couple of other sources of differences between the two readings. What if the lab meter isn't temperature compensated or is manually adjusted to an incorrect sample temperature? Was the sample taken from the exact same location as that of the measuring electrode of the online sensor?
HOW MANY SENSORS?
A diligent process designer may install a second pH sensor into the process, so the plant can average the two readings and use that average for the pH measurement.
Sounds good on paper, but in practice this isn't a reliable method. Why? First, it's important to understand that no two pH sensors are alike. In buffers, maybe — however, once the sensors contact the process, each will have a different take on what the process pH represents.
Perhaps a contaminant has hit one sensor, shifting its pH reading and thus affecting the average of the two sensors. Therefore, the overall average doesn't reflect the true pH of the process.
Or the sensors may be seeing different pH because of where they're mounted within the process. If the sensors are installed in a tank, maybe one is closer to a mixer or a reagent addition line. In either case, the two sensors aren't seeing the same value. Streaming potentials (an electric current that occurs when an electrolyte is driven through the sensor) also may affect the pH value, depending upon where the sensor is mounted in relation to the other sensor. Bottom line: Having two sensors is no better than having a single sensor.
The best option is to use three sensors and disregard the reading that differs most and average the other two to get a reasonable interpretation of true process pH. The sensors should be mounted as close as possible to each other to minimize potential differences.
Three sensors will cost a little more upfront. However, in the long run, the accurate measurements achieved can lead to substantial economic benefits in many applications — for instance, in any process where tight pH control will save reagents, increase product quality, prevent process upsets or equipment damage, or where fines may be levied if pH outfall exceeds regulatory limits.
And now, new devices like multichannel transmitters (Figure 2) with three-input capability and digital sensor technology reduce costs. You don't need three pH transmitters — just three pH sensors and one transmitter. The multichannel transmitter detects differences among the three pH sensors and calculates an average value of the two closest readings, which it then sends to the control system.
By using three pH probes and a single transmitter, you can shorten working time in highly critical applications.
A CLEAN SENSOR IS CRUCIAL
For a pH sensor to maintain accuracy, it must remain clean. The glass measuring electrode can't become coated; similarly, the reference electrode assembly can't become coated, plugged or otherwise contaminated by the process solution. I've provided guidance on proper cleaning procedures in a previous article, "Maintain Accurate pH Readings."
With a three-sensor system, it's easier to determine when a sensor needs cleaning. If one consistently reads pH differently than the other two, then it probably requires cleaning and calibrating. Without such a system, the plant has no way of knowing when a sensor needs service — so it typically performs cleaning and calibration on a scheduled basis, in some cases when the sensor is perfectly fine. This is a waste of time.
With a modern pH sensor and a multichannel transmitter, it's easy to replace the questionable sensor. It's simply a matter of swapping in a precalibrated unit.
FREDERICK KOHLMANN is analytical product business manager for Endress+Hauser, Greenwood, Ind. E-mail him at Fred.Kohlmann@us.endress.com.