Calibration & the problems it poses

• Supersaturation

• Initial DO > 9.0 mg/L
• Initial DO > saturation point (based on Temp & Pressure)

• Calibration approach

• Winkler – reagent quality
• Probe – (Air-saturated)Water or (Water-saturated)Air calibration?

• Blanks and the Calibration

• Blanks can --and typically DO-- fail simply due to calibration problems

Back to Top

A Word About Winkler...

Winkler remains the "gold" standard of calibration techniques. Winkler is based on chemistry, which means, as long as your reagents are fresh and prepared properly and accurately, your calibration will be accurate. We DO see differences when Winkler is compared to Probe technology...but the differences are in absolute DO measurements; critical QC results are the same.

Winkler was, at one point, THE best available calibration technique. The reality is, however, that advances in probe technology combined with the simple fact that Winkler calibration is time and labor intensive, are slowly contributing to Winkler's demise. Certainly, if Winkler works for you, and you have the time required, by all means continue to use it. Alternatively, rest assured that you CAN obtain consistent and reliable calibrations using a DO probe.

The keys to successful Winkler calibration are:

• Use air-saturated dilution water
• Use fresh reagents
• Standardize titrant
• Perform Winkler titration
• Check Winkler result against theoretical saturation
• Set meter
• Document your procedure!

Back to Top

Apples to Apples? or Apples to Oranges?

With air-saturated WATER calibration, you are treating samples and calibration alike... the probe in both cases is submersed in sample (or dilution water).

With water-saturated AIR calibration, samples and calibration are treated differently...the probe is in air for calibration, but submersed for sample analysis.

For many years, we have referred to AIR calibration as a case of "apples to oranges" since the probe sits in air for calibration but is immersed in sample (or dilution water) during sample analysis. This is distinctly different from the WATER (saturated air) calibration in which the probe is always immersed in sample (or dilution water). Historically, this difference has been the root cause of many a blank failure.

At the heart of this issue is physics. We know we can easily saturate water with oxygen by simply shaking the water in an environment that offers sufficient exposure to oxygen in the air. This is best accomplished using a large container which is half full of dilution water, offering sufficient headspace for air (and thus oxygen). Maintaining a water-saturated air environment, however, is a far more difficult proposition.

Consider an air calibration performed in the winter. It's quite possible that the atmosphere in a BOD bottle containing a little water, and sealed with the probe, represents 100% saturation (100% relative humidity). This is usually evidenced by the formation of condensation droplets on the probe tip. What's the first thing you do? Remove the probe from the bottle and either shake or wipe off the probe tip to remove the drops...right? Of course...but what have you done to the water saturated air in the bottle? If your lab is like most, the room air likely has a humidity of about 50% (or less). Removing the probe from the bottle causes a draft effect, which pulls the water-saturated air out of the bottle...to be replaced by air of much lower relative humidity. Now comes the physics. So you seal up the bottle again by replacing the probe. Physics pop quiz time: How longs does it take for the air in the bottle to become saturated with waster again. Here's a tip...it doesn't occur in just a few minutes. Frankly, this "passive saturation" can take hours, depending on barometric pressure, the room temperature, the amount of water in the bottle, and the amount of headspace of the bottle. Bottom line: There are just too many variables, and that's exactly why your blanks can deplete 0.3 mg/L one day and "create" 0.3 mg/L the next day. That's why we haven't been huge fans of AIR calibration. But...things are changing; technology is improving.

Back to Top

WATER vs. AIR calibration using the same sample

WATER Calibration example.

AIR Calibration example.

Chemists at the State Lab of Hygiene performed AIR and WATER calibrations in succession with the same sample, DO meter and probe. They noted that the AIR calibration result was higher by 0.15 mg/L (8.68 vs. 8.53) . NOTE that 8.68 divided by 1.023 equals 8.48, which represents the true saturation of oxygen at 742 mm pressure and 22.2° C. The factor 1.023 (102.3%) represents a correction factor used to adjust results obtained following AIR calibration to account for difference in temperature equilibration time. The calibration for new technology (dual thermistor) probes is 102.3% and 101.7% for older polarographic probes.

NOTE: These observations come specifically from one specific DO meter manufacturer and model number.

Back to Top

Common Sources of Error in the Water Saturated AIR Calibration

• Temperature: thermistor must equilibrate to air temp. in the BOD bottle before calibration

It takes time for the probe equilibrate initially after removing the droplets from the tip. It also takes longer to equilibrate in air than water! That’s why some have more success with the air-saturated WATER approach.

• Not allowing the air to become saturated with water…it takes at least 30 minutes.
• Not allowing the meter and probe to warm-up for at least 30 minutes
• Probe maintenance
• Change the membrane regularly
• Remove sulfide deposits from anode & cathode
• Check and calibrate the on-board barometer

Back to Top

Key to successful water saturated air calibration

• Consistency is the key
• Meter warm up time (at least 30 mins)
• How droplets removed from the probe tip (Shake? Dab?)
• Amount of water in the BOD bottle (~1 inch)
• How long you let the probe sit in the BOD bottle or the calibration chamber before calibration ( > 30 mins.)
• Consistent temperature conditions in Lab
• MUST be consistent from day 0 to day 5
• Get into a routine and STICK WITH IT!
• How important is consistency?

Your calibrations will likely work even if you don’t wait for the air to be 100% saturated with water as long as you do your calibration the same EVERYDAY!

One operator’s SOP for consistency:

• 30 minute warm-up for meter
• Allow 1-hour in bottle after wiping probe tip
• New membrane every 2 weeks
• Result: Has successfully met blank requirements for several years

Source: Joe Flannigan, Blanchardville Wastewater Treatment Plant; former Lab-of-the-Year Winner.

Conclusions: AIR vs. WATER Calibration

• There IS a difference when calibrating in air vs. measuring samples in water!
• Manufacturers often program a correction factor to account for the difference between oxygen diffusion in air vs. water.
• This is often seen as a saturation of 102.3% (polarographic) when calibrating in air vs. calibrating in water.
• Which technique is best?????? WHATEVER WORKS!

Back to Top

Correct Calibration Protocols

Water-Saturated AIR

• Place the probe in a BOD bottle containg about 3 cm of water.
• Shake BOD bottle prior to inserting probe to assure saturation. We recommend leaving the stirrer on (although manufacturer says it's not necessary) ---it speeds up equilibration.
• The probe may need to sit in the bottle for 30-35 minutes in order to match the temperature of the air.
• Determine barometric pressure and adjust meter's internal barometer as necessary.
• Check the temperature of the air (in the bottle) to be sure the probe thermistor is working correctly.
• Use the meter's auto-calibration function to calibrate the probe and meter.

Air-Saturated WATER

• Place the probe in a BOD bottle filled with air-saturated (well-shaken) water.
• Leave probe in the water w/ stirrer operating long enough for the probe temperature to equalize with the water temperature.
• Determine barometric pressure and adjust meter’s internal barometer as necessary.
• Check the temperature of source water to be sure the probe thermistor is working correctly.
• Use a detailed DO saturation table to determine the theoretical DO concentration.
• Adjust the meter to read the DO concentration determined from the saturation table.

Back to Top

Summary of Calibration Techniques

• The Winkler calibration takes longer than the other calibration techniques...with no net gain in quality.
• Calibration with air-saturated water takes less time because the probe’s temperature equilibrates quicker in water than air. This is because water is a more effective heat sink than air.
• You don’t have to worry about dealing with droplets on the probe tip when calibrating in air-saturated water!!!
• All three methods work. The results of the seed control and GGA were the same even though the IDO’s and DO5’s were different.
• Consistency is the key to good results regardless of calibration technique.
• Which technique is best?????? WHATEVER WORKS FOR YOU!

Back to Top

How do you KNOW DO measurements are accurate?

Download DO Saturation Table

The answer is: You need a "known" standard. If you have a standard which you know to be a specified dissolved oxygen concentration, and you measure that standard and obtain a reading equal to (or acceptably close to) that true value, then you have documentation to support the ability to obtain accurate measurements.

We also understand that labs have a lot of QC responsibilities already, so the best part of all this is that you can analyze a known standard without having to add any additional analyses! That's because, the initial DO reading of your calibration blank IS a known standard.

The process can be summarized as:

• Prepare Air-saturated (by shaking) dilution water
• Get some basic physical data
• Temperature

Remember: Temperature ± 0.5 °C translates to ± 0.1 mg/L DO

• Absolute barometric pressure
• Check reading against local TV station, radio, or airport at least monthly
• Compare apples to apples (Has the pressure value been corrected to sea level? (Most are). If so, you'll need to "UNcorrect" it.)
• Make adjustments as necessary

Remember: Pressure ± 5 mm translates to ± 0.06 mg/L DO

• Use physical data to determine standard "true" value (use a DO Saturation table)
• Determine theoretical TRUE value for oxygen in mg/L (i.e. saturation point)
• If measured value = True value ± 0.2 mg/L; calibration is accurate

Back to Top