DO Sensor Calibration

Not every dissolved oxygen sensor supplies the same output for a given oxygen concentration. Even for a given type of sensor (polarographic or optical), each sensor differs slightly in its physical characteristics, material composition, and membrane fabrication. For polarographic sensors, aging of the electrolyte causes the measuring current to decrease slowly with time.

Together these differences can result in slight variations of the sensor output from one sensor to the next. Therefore, each sensor has its own characteristic curve (called slope) of output current vs. oxygen concentration. To retain high measurement accuracy, the output of the transmitter must be adapted (calibrated) to the sensor’s zero-current and slope at regular intervals.

Sensor Zero Calibration

The zero current of the oxygen sensor at 0% oxygen concentration must be calibrated first. Normally the zero current is negligibly small and almost identical with the transmitter zero point. Nevertheless, the sensor zero current should be periodically checked, especially before the measurement of low oxygen concentrations.

Zero point calibration may be performed either in pure nitrogen or in water saturated with nitrogen. Another alternative is the use of a freshly-prepared 2% sodium sulfite solution. In any case, the sensor tip must be exposed to 0% oxygen.

The saturation of water with nitrogen can be time consuming, therefore calibration with pure nitrogen gas is faster and more reliable. The nitrogen gas must flow over the sensor tip. The zero point can be read after approximately 2–3 minutes.

The zero point calibration may often be omitted because modern oxygen sensors are “zero current free,” but it should be performed at least at the first-time calibration. Because today’s DO transmitters are mostly software-based, they can store the oxygen sensor data as reference values for sensor statistics. These are the deviations of slope and zero point, corresponding calibration values for temperature, air pressure, response time, and calibration date and time, all with respect to the reference values of the first-time calibration.

For polarographic sensors, a first-time calibration must be performed any time the complete sensor, the electrolyte, or the membrane with electrolyte is replaced.

(For more details on zero point calibration refer to the Best Practices article).

Together, these differences can result in slight variations of the sensor output from one sensor to the next. Therefore, each sensor has its own characteristic curve (called slope) of output current vs. oxygen concentration. To retain high measurement accuracy, the output of the transmitter must be adapted (calibrated) to the sensor’s zero current and slope at regular intervals.

Sensor Slope Calibration

The slope calibration can be performed in water or air where the oxygen concentration is known.

Calibration in Water

Water calibration is usually performed in 100% air-saturated water, i.e., the water must be in equilibrium with air. Oxygen exchange between water and air is normally quite slow. In order to accelerate the saturation process, air may be forced into water by a small fish-tank pump or with the aid of a compressed air bottle. Care must be taken that the air inlet is not placed too deep below the water surface and the air velocity is not too fast; otherwise, the air might be compressed by the water column, causing bubble formation. The bubbles would cause a higher partial oxygen pressure in the water than that above the surface.

Another requirement is an absolute constant water temperature during the calibration process. Water temperature might be influenced by changing room temperature and by the cooling effect of the forced air intake. Stabilizing the temperature can be a long process. As a result, it is quite difficult to produce air-saturated water with an accuracy better than 1 to 2%.

Modern DO transmitters include a small barometric pressure sensor. As explained previously, the oxygen sensor incorporates a temperature probe. Pressure and temperature are, therefore, always known, and the software automatically performs the necessary calculations for ambient pressure and temperature compensation for membrane permeability and water solubility.

For polarographic sensors, the calibration water must be stirred to achieve the minimum required flow velocity, and the water temperature and pressure conditions should correspond as closely as possible to those expected to be encountered during the actual measurement.

Calibration in Air

Because it is easier to produce water vapor-saturated air than air-saturated water, slope calibration should be performed in air. The partial pressure of oxygen, as explained previously, is the same above the water surface as that just under the water surface. A water bottle filled partly with water and left alone for approximately 15 minutes provides 100% water vapor-saturated air. It is important to keep the tip of the oxygen sensor dry because condensed water droplets at the membrane can influence the oxygen diffusion through the membrane and cause an incorrect electron current.

Parameters such as temperature and pressure must remain constant during calibration.

With modern DO measurement software such as ArcAir®, it is no longer required to perform the air calibration in water vapor-saturated air. An input is provided to manually enter the ambient relative air humidity in percent. The calibration is set to a 100% saturation index. The software calculates the partial pressure of oxygen at the set relative humidity and adds the missing partial pressure of the water vapor. The result is an indication of 100% saturation index either just above a water surface or just under the water surface. The measuring result in ambient air will be higher than 100% because the partial pressure of oxygen in ambient air will be higher because of the missing water vapor.

Hamilton ArcAir® software allows humidity to be input prior to calibration.

Hamilton ArcAir® Software allows humidity to be input prior to calibration. The partial pressure of oxygen in partially water vapor-saturated air can be calculated by:

Where:

P”_O2 = pressure of the partially water vapor-saturated air (kPa)
X_O2 = volume fraction of oxygen in air (0.2095)
P_air = total atmospheric pressure (kPa)
P_water = partial water vapor pressure (kPa)
h = relative humidity in %/100

Example: What is the saturation index indication of a DO meter in ambient air at 20°C, 101.3 kPa atmospheric pressure, and a relative humidity of 50%?

Note from our prior article, 20.73 kPa was the calculated value for the partial pressure of oxygen in water vapor-saturated air at equal conditions. The percentage difference can then be calculated as follows:

Influence of Salinity

If a Winkler titration is performed on air-saturated freshwater and, for the sake of comparison, on air-saturated seawater, both under identical temperature and barometric pressure conditions, the oxygen concentration (mg/l) in seawater will be observed to be approximately 20% lower than it is in freshwater. If these measurements are repeated with a Clark sensor, both measurements will render identical results, because the membrane-covered oxygen sensor responds only to the partial pressure of oxygen.

The partial pressure of oxygen in both freshwater and seawater under identical conditions is the same because it is governed only by the air above the water surface. It follows that saltwater does not require as much oxygen to establish equilibrium in the partial pressure of oxygen at the air/water boundary. The solubility of oxygen in water is reduced by the salt content in that water, and a dissolved oxygen measurement with a Clark sensor does not detect this fact.

An accurate dissolved oxygen concentration measurement with a Clark sensor is possible if one can refer to a special saturation oxygen table for saltwater. Some DO measurement software integrates these so-called oceanographic tables, supplied by Unesco, in order to calculate the correct oxygen concentration in seawater.

It is not possible to provide general guidance for the salinity correction when measuring the dissolved oxygen concentration in aqueous solutions. This depends mainly on the ability of the available measuring system; it is necessary to refer to the operating instructions.