The COD (Chemical Oxygen Demand) is one of most important parameters in water analysis. In Germany as in nearly all European countries tax for indirect discharges is based on the COD. Using the oxygen consumption is reasonable in the case of tax because the delay time of waste water in a treatment plant depends mainly on its biological activity. The oxygen consumption again is a degree of biological activity. To determine the biological oxygen demand would be best suitable, but determination of this parameter is even more extensive compared to measurement of COD.

COD (Chemical Oxygen Demand)

However, the disadvantages of COD have to be mentioned: the determination of the COD is circumstantial and requires a lot of time and manpower. The analytical procedure could be described as follows 1 : the sample has to be mixed with mercury sulphate, silver sulphate and potassium dichromate (if volatile compounds can be expected, an ice bed is necessary for the mixture). The mixture has to be boiled for 2 hours. Afterwards the amount of the remaining Cr(VI) has to be measured by titration with a Fe(II)-solution.

It should be mentioned that a lot of hazardous waste is produced during the analysis (in some rural areas in Germany water labs became the biggest producers of hazardous waste after the COD measurement became mandatory) and furthermore the digestion procedure isn’t that reliable at all. A lot of organic compounds cannot be oxidised by potassium chromate like organosilicium compounds.

The COD as described above cannot be used for an online parameter caused by the time-consuming process and technical equipment which is needed for the implementation in an online monitoring system. Nevertheless there are some online methods claiming to measure at least some kind of COD, where the samples were oxidised by different agents like compounds containing OH radicals or ozone.

These methods are suitable for sample matrixes of simple composition. Higher concentrations of salt or particles put the correlation of these alternative COD methods with the classical COD into jeopardy. For example the correlation between an ozone oxidation and the classical COD is shown in fig. 1. It is quite obvious that for filtered samples a correlation can be observed, but for unfiltered samples the results strongly differs

TOC (Total Organic Carbon)

Problems in practical lab work and online monitoring resulted in attempts to replace the COD by the TOC (Total Organic Carbon). The replacement is problematic because COD and TOC describe different parameters: to determinate the COD the oxygen consumption of a sample is measured whereas the TOC describes the concentration of carbon (neither COD nor TOC is a sum parameter). The TOC is measured by oxidation of the sample followed by IR detection of the CO 2 produced during the oxidation process. The oxidation can be caused by high temperature digestion or by UV oxidation, supported by oxidation agents like peroxodisulphate. For waste water analysis high temperature oxidation is appropriate since UV oxidation is unable to oxidise the TOC bonds in particles described elsewhere 2 .

A comparison of methods

In 2006 the replacement of COD by TOC was adopted in Germany based on the Wasserabgabengesetz (german law) which regulates the taxes for indirect dischargers but the realisation was not successful. It was pointed out that in the case of replacing COD by TOC costs for indirect dischargers will increase, because of varied correlation between TOC and COD3. Even within the same industry the correlation factor varied in a wide range:

  • Waste water in the power plant area 2.75
  • Oil business 1.5 – 7.0
  • Waste water soda lime industry 2.5
  • Pulp and paper industry 1.9 – 3.9
  • Textile finishing industry 2.2 – 6.6
  • Non-ferrous metal industry 1.1 – 8.8
  • Well 1.6 – 4.3
  • Sugar refinery industry 0.8 – 10.0
  • Chemical industry 2.0 – 6.8
  • Breweries 2.1 – 2.4
  • Fruit juice industry 2.7 – 2.9

Only for a few types of waste water a stable TOC/COD correlation can be defined, caused by the fact that the COD detects much more compounds than the TOC. In waste water analysis the COD is considered to be the sum of carbonaceous oxygen demand and nitrogenous oxygen demand 4 . The TOC considers the carbonaceous oxygen demand without consideration of the different oxidation states of the carbon compounds. For instance the oxygen demand of oxalic acid is much less than the oxygen demand or hydrocarbons because oxalic acid is already partly oxidised. Nevertheless 1 mmol/l oxalic acid resulted in the same value of TOC as 1 mmol/l ethane whereas the COD of oxalic acid is ~ 6 times lower. Therefore a simple replacement of the COD by the TOC is not possible. If the TOC has to be introduced in official legislation all the limit values must be re-evaluated in a time-consuming and high-cost process which will takes years, not to mention the costs. Furthermore the parameter COD is more interesting for waste water treatment plants because the organic matter in a sample is less interesting than the oxygen attrition of the sample.

Even if the samples are taken from the same spot in a very short time window (one day) the correlation COD/TOC is still unstable. The daily concentration curve of TOC and COD of the inlet water in a municipal 18 Water Monitoring & Treatment waste water treatment plant is given in fig. 2. The correlation factor of TOC and COD jumps between 3.5 and 6 during 24 hours.

References which suggest that the COD can be replaced by the TOC by applying a correlation factor are too optimistic in their hypothesis 5 .

TOD (Total Oxygen Demand)

Due to these setbacks linked to COD and TOC another parameter should be discussed: the TOD (Total Oxygen Demand). This parameter is simple to measure, does not require any chemicals and supports official institutions with almost the same information as the COD.

The TOD can be measured by high temperature oxidation of the sample under a defined oxygen atmosphere. The depletion of oxygen in the carrier gas is measured by a lambda sensor or another appropriate oxygen detector. The oxygen degradation correlates with the COD. A typical flowchart of a COD analyser is given in fig. 3.

In this example the sample is injected septumless via a ball valve directly into the hot furnace. Neither valves nor tubes are involved in the sample injection process, therefore the sampling takes place without particle discrimination. The furnace is working at a temperature of 1200°C. At this temperature the use of a catalyst is not necessary. The main advantages of abstaining of a catalyst is the ability of the combustion system to handle with salt or particle containing samples which may block or even poison the catalyst which leads to more maintenance efforts. Of course lower temperatures can also be used but according to experiences in high temperature TOC analysers a catalyst will be required. When using a catalyst it is important that the catalyst itself does not contain oxygen.

It is well known that at temperatures of >700°C ceramic materials like copper oxide, which is a common catalyst material in elemental and TOC analysis, releases oxygen, thus giving a false low reading. Furthermore catalysts may contain substances which can be oxidised, thus producing a blank problem.

TODth [mg/l] TODm [mg/l] RR [%] COD [mg/l] RR [%]
Acetaldehyde 1200 1130 94.5 796 66.4
Aniline 1560 1450 92.9 1020 65.6
1-Butanol 2310 2080 89.7 1860 80.3
Acetic acid 1210 1120 92.1 991 81.5
Ethanole 1940 1640 84.7 1280 66.0
Acetic ester 1680 1630 97.0 963 57.2
Glucose 2690 2670 99.3 2670 99.3
Methanole 1180 1180 100 918 77.5
Silanole 1690 1700 101 1060 62.6

For the oxygen detection a ZrO 2 sensor is used. The measuring range of this detector is 0-200,000 ppm O 2. Synthetic air with an oxygen concentration of 20% or nitrogen enriched by oxygen can be used for carrier gas. The latter is preferable if the TOD concentration of the sample is <20,000 ppm. The enrichment runs by a permeation unit which allows some oxygen from the ambient air to filter into the carrier gas stream. By this setup TOD concentrations from 10-200,000 mg/l can be measured.

The drastic digestion conditions of the TOD method always resulted in recovery rates of ~ 100% dissimilar to the COD. In table 1 the recovery rates of synthetic samples are displayed. The COD was measured according to ISO 8245, the TOD was measured following the method described above.

For all examples the recovery of the TOD is very close to the theoretical estimated value whereas the recovery rate of the COD depends strongly on the substances. It is of particular interest that compounds like acetic, ester or acetaldehyde which are easily biodegradable shows a rather poor COD recovery. On the other hand the TOD method covers substances which have the reputation to be very stable like silanole.

More interesting than these rather theoretical substances is the comparison of TOD and COD using real samples. Long term observations were carried out with waste water of a pulp and paper factory and with waste water of an industrial waste water treatment plant. It turned out that the correlation of these two parameters is excellent (fig. 4 and fig. 5).

In both examples the COD was analysed by the cuvette method where the COD is measured by mixing the sample with the chromate in a cuvette followed by photometric detection of the remaining chromate. Obviously the TOD provides the manager of the waste water treatment plant with the same information as the COD in a couple of minutes instead of two hours.

It should be mentioned that due to the oxidation principle some substances can release oxygen, thus giving lower TOD readings. For example nitrate will be transformed to NO and oxygen. 400 mg/l nitrate will release 20 mg/l oxygen, thus lowering the TOD value for 20 mg/l. Of course such high nitrate concentrations are exceptional. Considering the more common relations TOD/nitrate it is not likely that this effect has a serious impact on the TOD results. Chlorine, which strongly influences the classical COD, does not interfere with the TOD determination at all.


In conclusion the measurement of the TOD is more advantageous in comparison to the COD measurement: no hazardous chemicals are necessary, the acidification of the sample (which is common in TOC analysis) can be avoided. For analysis only a few minutes are needed compared to more than two hours for the COD measurement. The TOD method can be implemented in an automatic system, therefore saving manpower. The regulations for waste water controlling does not need to be modified because basically the TOD leads to similar results as the COD. The fast reaction time of TOD analysis enables the processor of a waste water treatment plant to react very fast on changes of the input of oxygen consuming material as well as of the effluent, thus saving both, money and environment.


1 ASTM D1252: Standard test method for chemical oxygen demand (Dichromate oxygen demand) of water

2 ISO 8245: Water Quality-Guidelines for determination of Total Organic Carbon (TOC) and Dissolved Organic Carbon (DOC)

3 BDI II/4-042-41/10: Positionspapier zu den Entwürfen der Änderung des Abwasserabgabengesetzes und der Abwasserordnung

4 Clifford, McGaughey: Simultaneous Determination of Total Nitrogen and Total Oxygen Demand in Aqueous Samples, Anal. Chem., 1982, 54, p.1345-1350

5 Braun, Furtmann, Stock: Improvement of the waste discharge control by establishing the new methods TOC, TNb and Ptotal-ICP, Texte/ Umweltbundesamt ISSN 0722-186X

Published: 10th Jun 2007 in AWE International