Activated Sludge (ASI)

What's Wrong with BOD?

Parameters like biochemical oxygen demand (BOD), chemical oxygen demand (COD) and total or volatile solids (TSS or VSS) are traditionally used to characterize organic matter in wastewater. These parameters provide an indirect determination of the quantity of the organic matter present. COD and VSS are measures of the total concentration of organic matter and particulate organic matter respectively, while the empirical BOD test provides a partial estimate of biodegradable organic matter only.

None of these methods give detailed information on the composition of the fractions of biodegradable and unbiodegradable organic matter. Some information on the wastewater biodegradability can be gained comparing different measures, e.g. BOD and COD, where a high ratio of BOD to COD shows that the wastewater is relatively biodegradable whereas a low BOD to COD ratio indicates that the wastewater is either more slowly biodegraded or contains a significant fraction of unbiodegradable material.

Historically speaking . . . . .
In 1912 the Royal Commission on Sewage Disposal took the view that the five day Biochemical Oxygen Demand (BOD) test was the most reliable chemical index of river water quality. The BOD figures recommended by the committee became known as the ' Royal Commission river classification'. It is perhaps reasonable to note that the BOD test was not introduced to measure the strength of raw wastewater, rather the effect upon the receiving waters. Interestingly, the 5-day duration for BOD determination has no theoretical grounding but is based on historical convention (Tchobanoglous & Schoeder, 1985).

"In a report prepared by the Royal Commission on Sewage Disposal in the United Kingdom at the beginning of last century, it was recommended that a 5-day, 18.3°C, BOD value be used as a reference in Great Britain. These values were selected because British rivers do not have a flow time to the open sea greater than 5 days and average long-term summer temperatures do not exceed 18.3°C. The temperature has been rounded upward to 20°C, but the 5-day time period has become the universal scientific and legal reference."

What are we really trying to measure anyway?
Since BOD is measured on the effluent, perhaps it is only natural to see BOD removal as an important process parameter. But should BOD be measured on the influent as well to provide that information even though it is an inappropriate measurement? BOD is seen by many agencies as the most appropriate way of measuring the strength of carbonaceous load, even though BOD is essentially inferior to COD in almost all ways including accuracy and precision. BOD is further affected by nitrification and although it is often assumed that the 5 day test of raw sewage will not nitrify, absence of nitrification cannot be guaranteed. In any particular BOD analysis even a small amount of nitrification will markedly increase the result.

Figure 1. The BOD curve (green) showing the superimposed nitrogen oxidation curve (red).

Incorporation of all or some of the ammonia-N in a BOD measurement is a needless complication, as ammonia-N can be measured directly rather than indirectly as an oxygen demand. What is important is carbonaceous load and even though CBOD does not accurately measure that, the needless complication of oxidation of ammonia should be avoided.

Nitrification inhibiting chemicals may be added to repress the bacterial oxidation of ammonia, and isolate the carbonaceous BOD5 or CBOD5 although the influence of these chemicals on the conduct of the CBOD test have been called into question (Albertson 1995). In this article the BOD5 referred to is actually CBOD5 because carbonaceous oxygen demand is what is required. If nitrification inhibition is desired 2-chloro-6-(trichloro methyl) pyridine (TCMP) or allyl thiourea (ATU) may be added at appropriate concentrations.

Why not use CBOD5 to measure wastewater strength?
CBOD5 will only measure a portion of the biodegradable matter in a wastewater sample, the question is how big is that portion? Wastewater contains both biodegradable and non-biodegradable material and what is often poorly understood is the way in which CBOD5 provides biodegradability data for a wastewater. Biochemical oxidation is a slow process and theoretically takes an infinitely long time to approach completion.

Because the biodegradable material in the wastewater will oxidise at different rates, in the 5-day BOD test, the extent of bio-oxidation will usually range from 60 to 80 percent of the biodegradable material. Twenty days is considered, by convention, adequate time for a complete biochemical oxidation of biodegradable organic matter in a water sample, but a BOD20 or an ultimate BOD test (also termed BODU, BODtotal, BODinf or BOD∞) is impractical when data are needed to address immediate problems.

Biochemical characterisation: Kinetics of BOD5
BCOD or biodegradable COD is the sum of the COD of both soluble and particulate biodegradable material (described in IWA activated sludge models as SS and XS respectively). There is also usually soluble and particulate unbiodegradable material present as well which are not detected in the BOD test (described in IWA ASM models as SI and XI respectively). Biodegradable COD (BCOD) fraction may be determined from a BOD analysis where the BOD is measured as a function of time (see below).

In general the BOD5 does not represent the total biodegradable fraction of the COD. Depending on the type of wastewater, anything between 50 to 95% of the biodegradable COD is oxidised after 5 days. After 20 days 95 to 99% of the biodegradable COD is oxidized but the measurement of the BOD20 is not reliable due to the extent of the dilutions involved and so not advised. A better approach is given by following the course of the BOD as function of time and calculating the BODU of the wastewater.

This can be achieved by setting up a number of replicated BOD tests which can be read at say 1, 2, 3, 4, 5, 6, 8 days respectively, providing BOD1, BOD2, BOD3 etc to BOD8. While this appears to be a lot of effort, it can easily be fitted into a BOD5 monitoring program. I personally prefer to use one of the manometric units with direct read facilities to a PC or some other device. This type of equipment allows you to read the same bottles at intervals over say 8-10 days (once the dilutions are worked out).

Without kBOD, BOD5 is not very useful. The idea that the BOD5 test provides valuable information in terms of biodegradability in isolation is spurious and one should not even compare BOD values unless the kBOD are of similar size in both samples of wastewater.

The kinetics BOD reaction are based on the classical first-order empirical rate equation:
BODt = BODu(1- e-kBOD.t).
where, BODt = BOD at t days, BODu = Ultimate BOD, kBOD = rate constant (d-1) and t = time (days)

The relationship between BODU and BOD5 can be expressed as follows:
For a BOD of t days, BODt = BODu(1- e-kBOD.t).
So, if the BOD5 of a wastewater is 280 mg L-1, and kBOD is 0.45 d-1:
BODu = 280/(1 � e-0.45x5) = 313 mg L-1

BODu is the total or ultimate oxygen biological demand and is approximates the Biodegradable COD (BCOD) content of the wastewater.

Without kBOD, BOD5 is not very useful. The idea that the BOD5 test provides valuable information in terms of biodegradability in isolation is spurious and one should not even compare BOD values unless the kBOD are of similar size in both samples of wastewater. Two or more samples may have a similar BOD5 value, but this may be due to a combination of different BODU and KBOD values as shown in Table 1.

Table 1. Comparison of a fixed BOD5 value and possible BODU, BCOD and kBOD combinations.

BOD5 mgL-1 BODu mgL-1 kBOD d-1 BCOD mgL-1
250 281 0.44 332
250 350 0.25 413
250 422 0.18 500

Fortunately, oxygen demand, whether biochemical or chemical, are essentially the same thing (oxygen used for oxidation of organics) with the same units (mgL-1 O2), and if we measure biodegradable COD we are also measuring all of the BOD. Total COD estimates all the BOD and nonbiodegradable COD fractions, and BOD estimates a proportion of COD, the amount of which depends on the size of kBOD.

Figure 2. Effect of kBOD on CBOD5 curves for a constant value of CBODu (300 mgL-1)

Now it is reasonable to accept that the CBOD5 for a particular sample is say 250 mg O2 L-1, but understanding precisely what that means is the issue as the rate dependence of the BOD test is its greatest problem, particularly when kBOD is low (see Table 1). The Glucose/Glutamic acid (GGA) BOD standardization which uses the 300 mgL-1 biodegradable mixed primary standard, should have an average BOD5 of 198 mg/L with a standard deviation of not greater than 30.5 mgL-1 (±15.4%). When nitrification inhibitors are used, GGA test results falling outside the 198 ± 30.5 mg control limit quite often indicate use of incorrect amounts of seed requiring an adjustment of the amount of seed added to the GGA test to achieve results falling within this range by either raising the rate or lowering it (APHA 1998). All other things being equal, the variations are essentially due to variation in the magnitude of kBOD.

. . . . it is reasonable to accept that the CBOD5 for a sample is say 250 mg O2 L-1, but understanding precisely what that means is the issue, as the rate dependence of the BOD test is its greatest problem, particularly when kBOD is low.

BCOD or biodegradable COD is the sum of both soluble and particulate material (described in IWA ASM models as SS and XS respectively). There is also usually soluble and particulate unbiodegradable material present as well which will not be detected in the BOD test (described in IWA ASM models as SI and XI respectively)

Table 2. Varying CBOD5 values for samples containing BODU of 300 mg L-1

KBOD d-1 CBOD5 mgL-1 % of BODU
0.10 118 39
0.15 158 53
0.20 190 63
0.30 233 78
0.40 259 87
0.60 285 95

The example shown in Fig. 1 and summarized in Table 2 presents CBOD5 data for 6 samples which all have the same CBODu (300 mg L-1 or 100%), but have significantly different KBOD values. As we progress from KBOD of 0.10 to 0.60 the overall biodegradability and the rate of oxidation of the wastewater increases, i.e. the proportion of slowly biodegradable material decreases while the readily biodegradable fraction increases.

This creates obvious probems in estimating BODu from BOD5 unless KBOD is known. BODu is often assumed to be 1.5 x BOD5 by some authors (Grady et al 1999) and 1.8 x BOD5 by others (Weijers, (1999). As rules of thumb go, this can be a very poor estimate (a BODU/BCOD5 factor of 1.5. represents a KBOD of about 0.22 d-1, while a factor of 1.8 corresponds to a KBOD of about 0.16 d-1 which effectively ignores the variable kinetic aspects of the test).

The use of BOD as a process control parameter
From what has been outlined above, it should be clear why BOD is potentially a poor means of measuring the carbonaceous load of raw wastewater. Never the less this test is still commonly applied to municipal activated sludge plant influents even though COD is a superior test and used by some municipal plants and a variety of industrial plants for many years. COD is becoming more commonly used in place of BOD in process monitoring and modeling. But often a simple conversion is made to convert BOD to COD or vice versa or both tests are applied. BOD/COD ratios should be used with care if one is attempting to calibrate a process model to a particular wastewater.

Similarly the use of BOD in establishing the F/M ratio at a plant renders the value of this parameter suspect. Any determination using BOD is essentially of historical interest only as it reflects samples taken a week previously, and MLSS and/or MLVSS are poor estimates of active biomass. It is true that F/M has been used in some plants for years and with reasonable success, but I would suggest that the values are meaningless in any situation except the plant in which they were determined and the data is again largely of historical interest rather than an important operations control parameter.

Readily and slowly biodegradable components of wastewater
The ratio SS/XS controls the value of kBOD. High concs of SS (SS/XS ≥1) will bring about a higher value of kBOD and the BOD5 test will more closely approximate BODu than it would if the concentration of XS was high (SS/XS<1).

A sample which contains a large amount of soluble readily biodegradable material and a small amount of slowly biodegradable/particulate material will have a high value of KBOD and BOD5 may be >90% of the CBODu, while a wastewater sample with a low RBCOD and high SBCOD may have a CBOD5 which is < 50% of CBODu.

The magnitude of kBOD depends on the relationship of SS and XS, which is influenced by the source and history of the wastewater, conditions in the reticulation system, the fraction and type of industrial wastewater and the application of pre-treatment.

The variability of wastewater strength and composition over a day (COD and SS/XS ratio) indicates that kBOD will vary significantly over the course of a day as well as from day to day. This indicates that flow weighted composite samples of wastewater should always be used, and that grab samples introduce the possibility of further uncertainty.

When the kBOD value is known (through measurement), the biodegradable COD (BCOD) can be determined from the CBOD5, which is often measured routinely. Because influent composition can vary over time, a verification of the kBOD value should be carried out regularly.

Conclusion.
The variability of wastewater is such that it is often difficult to reliably estimate any parameter not actually measured. While the BOD test is often used to measure the strength of wastewater there are a number of good reasons why this test provides potentially poorer data than the COD test. The kinetic basis of the BOD test is certainly a weakness as is uncertainty in the relationship between the BOD test kinetics and the biological processes in the plant. The CBOD test is certainly appropriate as a means of monitoring the biodegradability of plant discharges and this was always the intent of the test when it was introduced.

The BOD test also has the difficulty that the kinetics of the procedure is affected by the amounts of slowly biodegradable and readily biodegradable COD. The measurement of these parameters are of importance in the operation of plants which denitrify and/or remove phosphorus and if this aspect is addressed by determining the kBOD which then provides valuable information including BODU and BCOD. As KBOD increases, CBOD5 more closely approximates BODU and hence CBOD. The relevance of using kBOD, CBODU and BCOD to assist in the characterization of wastewater (fractionation of wastewater COD) is addressed in the second part of this article.

References and further reading

Albertson O.E. (1995) Is CBOD5 viable test for raw and settled wastewater. J Env. Eng pp 515-520
American Public Health Association, (1998) Method 5210 B, 5-Day BOD Test, "Standard Methods for the Analysis of Water and Wastewater", 20th edition, APHA, 1998
Bowman, G. and Mealy, R. "BOD 101: Everything You Always Wanted to Know About Biochemical Oxygen Demand (BOD) Testing but were Afraid to Ask." Training jointly sponsored by the Wisconsin Rural Water Association, the Wisconsin Department of Natural Resources-Laboratory Certification Program and the Wisconsin State Laboratory of Hygiene. October 13, 14, November 2, 4 and 8, 1999. Presented at various locations across Wisconsin. Availabe here. Biochemical Oxgyen Demand Training Course Handout (Fall 1999) (PDF file, 2,949KB)
Grady, C.P.L., Jr, Daigger, G.T. and Lim, H.C. (1999). Biological Wastewater Treatment. Second Edition,. Marcel Dekker, New York.
Tchobanoglous, George & Schoeder, Edward D., Water Quality Characteristics : Modeling & Modification, Reading, MA: Addison-Wesley Publishing Co/May, 1985
Stefan R. Weijers, (1999) On BOD tests for the determination of biodegradable COD for Calibrating Activated Sludge Model No. 1. Wat. Sci. Tech Vol 39 No 4 pp 177�184
On-line reference material. USGS BOD Field Manual