Water-Testing Interferences

Reader note: This article appeared in the AWT Magazine "The Analyst," as well as www.awt.org.

With so much at stake, it’s imperative that water treaters thoroughly understand the water chemistries in a treated system. We need to be the expert consultants, and proper water testing in the field is the foundation for becoming that expert. There are “interferences” which can keep a water treater from obtaining accurate results on the condition of the water in a system. Understanding these interferences will enable us to offer the best consultation.

During my presentation at the 2017 convention in Grand Rapids, I mentioned the fact that water treaters are often the largest source of interference. Poor techniques in sample collecting and testing methods can steer you away from the correct test results. Remember, meters need to be calibrated or proved accurate. Reagents need to be fresh. Show those reagents some respect; they don’t like cooking in the trunk of your car or being subjected to freezing temperatures. And if you refill your reagent bottles from larger bottles to save money, it’s a good practice to periodically replace the dropper tips. Finally, observe any waiting times required for a test method―they are there for a reason!

In addition to the problems we ourselves can bring to the testing arena, here are some common interferences you might encounter:

Sulfite tests: We’ve all collected hot samples out of a boiler’s sight glass or blowdown line. But did you know testing that sample while it is hot will give you a false-high sulfite reading? The sample must be cooled in a sealed bottle and allowed to reach room temperature before performing the test. Allowing to cool in an open container will result in a false-low reading.

Hardness tests: Metals like iron and copper interfere with the endpoint color development in hardness tests. The titrant for hardness tests is EDTA chelant. Chelants prefer metals more than they do calcium and magnesium hardness, so the chelant preferentially reacts with the iron and copper in a sample. The result…a “fading endpoint.” The sample will probably go from red to blue, but in a second or two it will turn back to red. Many additional drops of titrant will need to be added before the sample remains blue. To remove this interference, simply add two or three drops of titrant first, followed by the buffer and indicator; then do the titration. Remember to include those first two or three drops in the total when calculating the test results.

Ampule tests: The ampules used for testing residual oxygen in boiler feedwater samples can often give a false-high reading―but not because of their chemistry. This relates to letting a sample flow long enough to reach equilibrium while flushing out any oxygen in the sample tubing. The longer you let an unbroken ampule sit in the flowing sample, the more accurate the results will be.

Chlorine tests: High chlorine levels can bleach out the DPD indicator and prevent a pink color from developing. This usually occurs at chlorine levels over 10 ppm, which explains why we rarely see this in the industrial world. This bleaching interference could lead you to think there is not enough chlorine in the system, so you add more. A telltale sign that this is occurring is when you see a flash of pink in the sample when the DPD is added, but then the sample goes back to colorless. Dilution with a non-chlorinated water is necessary to determine the correct chlorine levels in these waters.

Total alkalinity tests: When a sample goes from blue to yellow instead of green to red after adding the total alkalinity reagent, this indicates an excess amount of chlorine is present in the sample. Although a yellow endpoint is still accurate, it is unexpected. To counter this unusual color change, add two drops of sodium thiosulfate to the sample before adding the indicator. The thiosulfate will remove the chlorine interference.

Phosphonate drop tests: Interferences in this type of test are frequent in the industry. First, phosphonate drop tests require the pH to be in the method’s spec range. The sample pH must be correct (2.6‒3.0). No one buffer handles all waters. Without confirming the sample has the correct pH level, the test result could be compromised. Second, fluoride is a positive interference with this test. Most municipal drinking-water systems add fluoride, which then cycles up in cooling waters. Without the addition of a fluoride masking agent, the test results will show higher levels than are actually in the cooling water. This will lead to undertreatment and poor results. Ortho- and polyphosphates are also positive interferences for this test method, while iron is a negative interference.

Iron and copper: These metals interfere with each other and with many other tests. If your corrosion rates are not under control, then your test results can lead you further from desired results.

“O” Alkalinity tests: Many people still use the “2P-M” method to determine “O” alkalinity in boiler waters. Polymers can titrate as “M” alkalinity, so high polymer levels can decrease your “O” alkalinity results – even making your “O” alkalinity measurement negative! For boiler waters using polymer, the “barium chloride” method for determining “O” alkalinity levels must be used.

Conclusion

Water treatment requires the water treater to be an expert on the water systems. There are mechanical and chemical aspects of these systems; water treaters need to understand both. The chemical side begins with having an accurate knowledge of the water chemistries and understanding testing interferences that can lead to the true conditions in the system.