Chapter 7 BOD Biochemical Oxygen Demand return to Table of Contents

The Biochemical Oxygen Demand test measures the ability of naturally occurring microorganisms to digest organic matter, usually in a 5 day incubation at 20º C, by analyzing the depletion of oxygen. This measures biodegradable organic matter. This is normally expressed as O₂ mg/L.

The biochemical oxygen demand analysis is an attempt to simulate the effect a waste will have on the dissolved oxygen of a stream by a laboratory test. The BOD test gives an indication of the amount of oxygen needed to stabilize or biologically oxidize the waste. The advantage of the BOD test is that it measures only the organics which are oxidized by the bacteria. The disadvantage is the 5 day time lag and the difficulty in obtaining consistent repetitive values. The COD test can be performed in a few hours, however, the results of the COD (chemical oxygen demand) tests are usually higher that the corresponding BOD test for several reasons. Many organic compounds which are dichromate oxidizable are not biochemically oxidizable; Certain inorganic substances, such as sulfides, sulfites, thiosulfates, nitrites and ferrous iron are oxidized by dichromate, creating an inorganic COD, which is misleading when estimating the organic content of the wastewater. The BOD results also may be affected by lack of seed acclimation, giving erroneously low readings while the COD results are independent of seed acclimation.

BOD values are a measure of food for naturally occurring microorganisms. When food is introduced, naturally occurring microorganisms begin to multiply rapidly at an exponential rate, resulting in the reduction of dissolved oxygen in the water. After all the food is consumed, they die and become food for protozoa, which in turn die off and become food for rotifers and crustaceans.

The Biochemical Oxygen Demand test provides a quantitative measure of the amount of oxygen required to maintain the growth and activities of the biological organisms responsible for the aerobic digestion of the organic and putrescible matter in the liquid at a set temperature for a set amount of time in the dark. There is a limited amount of oxygen which will dissolve in the sample (about 9 mg/L), therefore dilutions are prepared with a buffered mineral nutrient containing water. The test does not determine the total amount of organic material present, since many compounds are not decomposed by biological and biochemical reactions in the set conditions. The BOD indicates the concentration of biodegradable organic material. When applied to the examination of effluents, it permits computation of the efficiency of operation of the treatment units.

The DO is measured at the beginning and recorded. After five days the DO is again determined. The BOD is then calculated on the basis of reduction of DO and the size of the sample.

In conducting the test, several different dilutions are to be used to ensure that at least one will deplete the oxygen content by about 50%. For computation of the BOD value, at least 2 mg/L of O₂ should be consumed with at least 1 mg/L remaining. If this condition is not found in any of the dilutions used in the test, a wider range of dilutions must be selected for the examination of succeeding samples from this source. Two different dilutions of the same sample should generate valid results. The BOD₅ values of the two dilutions are normally averaged. If this is true, a BOD value for a glucose/glutamic acid standard should be 60-70% of the COD value for the same sample. Check a few old Water Pollution Studies ... you will find that BOD is ~64% of COD for the last umpteen studies. Ultimate Oxygen Demand is usually expressed as (1.5 X BOD₅) + (4.6 X NH₄-N). The 1.5 factor for BOD₅ accounts for the fact that only 60-70% of the BOD is oxidized in 5 days. To encourage the bacteria to consume ALL the oxidizeable material, the incubation would have to be allowed to go for 21 days or more.

As an aside, for some samples [those in which the organic material is completely biodegradable], the UBOD would be very close to a COD, since the COD also measures essentially all the oxidizeable material.

The minimum detection limit (MDL) is one thing that you can measure in the lab. There are different procedures for doing that, but one of the most commonly used is in Appendix B of Chapter 40, Code of Federal Regulations, Part 136 (40 CFR 136), the code (law) governing the NPDES program. Appendix B calls for preparing a sample of known concentration that is near but higher than the MDL mentioned in the written method. You analyze that sample seven times in one batch, calculate the standard deviation of the seven results, and multiply the standard deviation by 3.14. The answer is that lab's MDL for that method.

SM₁₈ 5210B states the MDL for BOD is 2.0 mg/L. That's as low as you can go because of the method requirement that a valid test must have a depletion of at least 2.0 mg/L. If the sample analyzed was say 295 mL sample and 5 mL dilution water (to get the needed nutrients [may not be enough]), the dilution factor would be essentially 1, and the BOD would therefore be 2.0 mg/L x 1 = 2.0 mg/L. That doesn't mean that the lab can actually analyze accurately at that level, though. To find out what your lab's actual MDL is, analyze a sample seven times, at a concentration of say 10 mg/L, just as called for in Appendix B, 40 CFR 136.

Matrix spikes are performed on reagent water rather than a real sample because of problems with interpretation and narrow acceptable dilution ranges. Acceptable values for the glucose-glutamic acid (G/GA) solution give 198 ± 30.5 mg/L BOD₅. The potassium hydrogen phthalate (KHP) solution has a BOD₅ of 240 mg/L, although many wastewater labs in Georgia report their mean around 260 mg/L. The obtained value will depend to a great extent on the mixed microbiological population. Control charts should be maintained for accuracy on either the G/GA or KHP test solutions and for precision on real sample duplicates.

Samples suitable for BOD₅ testing should be obtained prior to chlorination. When chlorinated effluent is tested it must be dechlorinated (sodium sulfite, Na₂SO₃) and seeded. If dechlorination is required, refer to the method for directions, as excess dechlorinating reagent will affect the BOD₅ result.

The BOD₅ calculation takes into account the initial and final DO values and the percent concentration. In the BOD test you will use the difference of 2 readings. The actual readings themselves could be off 0.5 mg/L and may not make any difference, as long as the calibration procedure [and everything else] is consistent. The Winkler titration has chemicals which are unstable (titrant) and goofy measurements (201 ml) and requires a very consistent technique. Air calibration requires a BOD bottle with an inch of distilled water, and a barometer if you want to be real accurate. There is less chance for error using the air calibration method, especially with a meter which is set up to do it automatically. The use of a barometer is crucial to achieving the < 0.2 mg/L change in the dilution water. When talking about the G/GA solution giving a result of 198 ± 30.5 mg/L, you should explain that this means the average should be in the vicinity of 198, and the standard deviation should be WELL BELOW 30.5, the "well below" being, as explained in the 19th Edition, because the 30.5 came from an interlab study involving many sources of variability not found in a single lab. If a lab's average is well above 198.... say 240 ... the regulatory agency shouldn't care, because all that means is that the samples being seeded by the lab (e.g., effluent samples) are being reported at, say, 20 mg/L when they are really lower ... a situation that protects the environment. If the lab's average is considerable LOWER than 198, just the opposite is true, and the regulatory agency should be on the lab's case. In the table of suggested dilutions for various BOD samples, at the bottom of the table where the suggested volumes to be diluted are 1.5 milliliters or smaller [ such volumes are very difficult to measure accurately] and the entire sample should be diluted before measuring the aliquot to incubate. Caution the analyst to remember to multiply the BOD result by that additional dilution factor. Source water is the single most common cause of bad blanks, and it deserves a discussion. Labs that use their own distilled or DI, or both, often have chlorine, ammonia, organics, and/or trace metals in the source water. All of these have a bad impact on BOD tests, and must be avoided in the source water. Sometimes it is best to buy distilled water (NOT distilled drinking water) ... steam distilled water seems to work best.

1. Background

a. The purpose of the biochemical oxygen demand (BOD₅) test is to measure the potential of wastewater and other waters to deplete the oxygen level of receiving waters. Wastewater treatment plant operators also use the BOD₅ test to determine the efficiency of their plant by measuring the BOD₅ of the influent, and of the effluent, and then calculating the percent removal. The BOD₅ test may also be used to test waters other than wastewaters to determine their oxygen-depleting potential.

b. Oxygen is required (used up, depleted, consumed) by microorganisms as they assimilate various organic and inorganic materials transported by the water. Some of those materials contain reduced forms of nitrogen. Collectively, these materials are called nitrogenous materials, and they are consumed by nitrifying bacteria. If those nitrifying bacteria are inhibited, usually by using a pyridine compound, the test measures only the carbonaceous organic material, and is referred to as carbonaceous BOD₅, or CBOD₅. Carbonaceous material is that consumable material which does not contain nitrogen.

c. While the Term "BOD" technically refers to the oxygen uptake demand of contaminated water, the term is commonly used to refer to the organic and inorganic materials consumed by bacteria, as if "BOD" is a contaminant in the water. For example, wastewater treatment plants report "percent removal of "BOD" as an indication of how well the plant is performing.

d. Many biological treatment plant effluents contain enough nitrifying bacteria that their consumption of nitrogenous material (material containing nitrogen) is significant. Because nitrogenous demand has historically been considered an interference (i.e., the primary purpose of the BOD₅ test is to measure carbonaceous material ... nitrogenous material can be measured by other tests such as ammonia, nitrate-nitrite, and total Kjeldahl nitrogen), the CBOD₅ test is a measure of only the carbonaceous material, one would expect a correlation for BOD₅ and CBOD₅ for a given sample.

e. It would take the bacteria in a wastewater sample 20 - plus days to assimilate all the consumable material in a water sample. Because it is impractical to spend 20 days waiting for the test to finalize, a 5-day test has been established as the standard (hence, BOD₅). Depending on the specific waste stream, 60 to 70 % of the available material is normally consumed in a 5-day period. A reasonable compromise. Another test that measures the oxygen requirements of wastewater is chemical oxygen demand, COD, which can measure 100 % of the chemically oxidizable material in less than 3 hours. One can expect BOD₅ and COD results for the same sample to be related. Normally, the BOD₅ value will be 60 to 70 % of the COD test result. Also, because BOD₅ is a measure of nitrogenous and carbonaceous material, and CBOD₅ is a measure of only the carbonaceous material, one would expect a correlation for BOD₅ and CBOD₅ for a given sample.

f. The BOD₅ test is widely used to determine the degree to which a waste stream will contribute to pollution of receiving waters by depriving organisms in those waters (fish) of their source of oxygen. The BOD₅ test is of prime importance in regulatory programs and in determining the overall health of receiving waters.

2. Approved Methods

a. The BOD₅ test is normally required by a regulatory program that is governed directly or indirectly by Chapter 40, Code of Federal Regulations, Part 136 (40 CFR 136). Currently, 40 CFR 136 allows the use of four analytical methods for testing and reporting BOD₅ and/or CBOD₅. EPA Method 405.1, Standard Methods 5210B, USGS Method I-1578, and AOAC Method 973.44 are the four allowed methods. EPA Method 405.1 refers the user to Standard Methods 5210B for specifics on conducting the test. It is Standard Methods 5210B or its derivative that is most commonly used in environmental laboratories.

b. All BOD₅ methods approved by EPA are based on the determination of oxygen depleted by bacteria in samples during a 5-day period, in the dark, at 20 ± 1º C as those bacteria consume organic materials in water samples. BOD₅ samples are normally incubated in 300 mL BOD bottles, and are diluted as necessary to allow at least 1.0 mg/L of DO (dissolved oxygen) to remain at the end of the 5-day period. To be a valid test, at least 2.0 mg/L of oxygen must be depleted during the incubation. If necessary, samples must be pretreated to assure proper pH, temperature, and absence of toxic materials (e.g., chlorine) thus creating a suitable environment for survival of the BOD-consuming bacteria. The foregoing are requirements of all EPA-approved BOD₅ methods.

3. Sampling, Sample Preservation, Holding times

a. Sampling Location. Discharge permits or other regulatory documents usually specify sampling locations. Care must be taken to make sure the sample is representative of the water body from which the sample was taken.

b. Sample Size. Samples should be taken in a clean plastic or glass container of sufficient size to provide enough sample for all of the tests and the quantity of BOD bottles that will be incubated for that sampling site. Another consideration in determining sample size, is that the sample must be representative of the waste stream, and the smaller the sample, the more difficult it is to make sure that it is representative.

c. Preservation and Holding Times

(1) In most environmental samples, bacteria naturally present in a waste stream are consuming oxygen before, during and after the sample is taken. Therefore, starting the analysis as soon after taking the sample is very important. 40 CFR 136 establishes the maximum storage (holding) time as 48 hours. Standard Methods 5210B recommends that the holding time not exceed six (6) hours, and if analysis cannot start within 6 hours, the sample may be held at 4º C for up to 24 hours (not freezing). Standard Methods 5210B also says that if analysis can start on a grab sample within two hours of sampling, preservation at 4º C is not necessary. Composite samples must be kept at 4º C during the entire compositing period, which should not exceed 24 hours. After removal from the compositor, samples must be stored at 4º C if analysis cannot start within two hours, and, just like grab samples, Standard Methods 5210B says they should not be held longer than six hours after removal from the compositor. Discharge permits specify whether samples should be grab or composite.

(2) Some labs operated by permitted wastewater dischargers may have difficulty meeting even the 48-hour holding time requirement, especially when required to do BOD₅ testing each and every day of the week. A variance is allowed by 40 CFR 136 if the permittee has data showing specific types of samples under study are stable for a longer time (unlikely to add more than 24 hours [72 hours total]).

4. Apparatus

a. Measurement of Dissolved Oxygen (DO). 40 CFR 136 allows measurement of DO using either an electrode and meter, or the azide modification of a Winkler titration. Applicable methods are EPA Method 360.1 or Standard Methods 4500-O G (electrode), or 360.2/4500-O C (Winkler).

(1) Electrode. An oxygen-sensitive membrane electrode, polarographic or galvanic, with appropriate meter meets EPA-approved method requirements. The YSI 50-series, and Orion 800-series instruments are examples of commercial meters meeting those requirements for measuring DO. The method calls for calibrating the meter using oxygen-saturated water, water-saturated air, or a Winkler titration before each use. If DO is routinely measured using an electrode and meter, checking calibration of the meter periodically with a Winkler titration will help make sure the meter is functioning properly. Standard Methods 4500-O G is a commonly-used electrode method for measuring DO.

(2) Winkler. Burettes capable of measuring accurately to 0.1 mL are sufficient for the Winkler titration. Standard Methods 4500-O D is a commonly-used method for measurement of DO using a modified Winkler titration.

b. Thermometers. Temperature in the BOD incubator is measured to ± 1º C accuracy using a thermometer that is either NIST-certified (National Institute of Standards and Technology, formerly NBS [National Bureau of Standards]), or has a calibration certificate or other documented evidence showing it is traceable to a NIST-certified thermometer. If using an air incubator, the thermometer should be immersed to the immersion line in a suitable container (e.g., beaker) filled with water or glycol solution which acts as a heat sink. If using a water bath, simply immerse the thermometer to the immersion line. Calibrate a less expensive thermometer for general use.

c. Incubator or Water Bath. Either an air incubator or water bath may be used to incubate BOD bottles. The apparatus must be able to maintain a temperature of 20 ± 1º C for the entire 5-day incubation period. It must be of sufficient size to hold all BOD bottles for a given batch (i.e., environmental samples, blanks, seed controls, glucose/glutamic acid standards). For reasons explained later, it is also advantageous for an air incubator to be large enough to hold the dilution water container used for the BOD₅ determination. The incubator/water bath must exclude all light to prevent photosynthesis (which would result in a positive contribution to dissolved oxygen in the incubation bottles, resulting in a negative bias in the BOD test). Incubators specifically designed for BOD₅ (such as Hach Company's Incutrol®) are available from scientific supply vendors. A household refrigerator can be modified to meet BOD₅ test needs. One required modification is installation of a small fan to create an airflow and ensure an even temperature throughout the refrigerator. The thermometer used to monitor temperature in the incubator should be placed in the vicinity of the majority of the BOD bottles.

d. Dilution water container. A glass or plastic container of laboratory grade (e.g., not an old milk bottle), large enough to supply all dilution water to be used in a given batch of samples should be used as the dilution water container. Although not required, a convenient way to introduce dilution water into BOD bottles without creating air bubbles is to siphon the water from the dilution water container. If this technique is used, the siphon hose should terminate with a 6-inch length of glass tube for filling the BOD bottle from the bottom without submerging the hose. The dilution water container and all associated equipment must be kept clean (washing with detergent and rinsing with distilled water should be sufficient).

e. Incubation (BOD) Bottles. Bottles used for BOD₅ tests should never be used for any test other than BOD₅. They can be either 75-, 250-, or 300-mL (300-mL is the most widely used). Just like the dilution water container, BOD bottles must be kept clean. Wash after each use with detergent, rinse with distilled water, drain, and store such that the bottles are not exposed to dust or other contaminants in the lab. If blanks unexpectedly test higher than normal, it may mean the BOD bottles need acid washing. This can be done by first washing, rinsing, and draining as above, and then rinsing with 1:1 mineral acid (purchased ready-made or prepared by slowly adding concentrated hydrochloric or nitric acid to an equal volume of distilled water). Rinsing is accomplished by carefully swirling 10 to 20 mL of the dilute acid until all inner surfaces of the bottle are wetted. Allow the bottle to sit for a few minutes before properly discarding the acid. Then rinse the bottle with distilled water and drain. Acid rinsing is not necessary every time the BOD bottles are used and might never be required as long as blanks test less than 0.1 mg/L. BOD bottles used for blanks should be chosen randomly to avoid checking only the "cleanest" bottle.

5. Reagents

a. Buffer. To provide the optimum environment for survival of bacteria in the incubated sample, it is necessary to buffer the sample such that it maintains a pH of 6.5 to 7.5 SU. The buffer can be prepared with various phosphate compounds, or it can be purchased ready-made. If stored, the buffer solution should be refrigerated at 4º C to preclude biological growth. It must be discarded if such growth appears because the growth has an oxygen demand which would introduce a positive bias into all BOD₅ measurements. The pH of the buffer solution should be 7.2 SU.

b. Nutrients. In addition to the nutrient value of the phosphate buffer, nutrients in the form of ammonium chloride, and trace metals in the form of ferric chloride, magnesium sulfate, and calcium chloride are added to the dilution water. These solutions can be purchased ready made (all combined in one packet), or they can be prepared individually. Hach Company, North Central Labs, and perhaps others sell packets (pillows) containing both buffers and nutrients.

c. Standards. To provide a check on efficiency of the seed and effectiveness of dilution water, Standard Methods 5210B implies that a standard solution should be analyzed with each batch of BOD₅ samples. The most common standard is a solution containing 150 mg/L glucose (dextrose) and 150 mg/L of glutamic acid, commonly referred to as the "G/GA" solution. This solution can be prepared as described in Appendix A, or purchased as a solution from a commercial vendor such as Hach Company or North Central Labs. In an interlaboratory study, several labs reported results for the G/GA test that averaged 198 mg/L BOD₅, with a interlaboratory standard deviation of 30.5 mg/L. As discussed in Standard Methods, a single lab should expect to achieve an average over several tests of the G/GA standard in the vicinity of 198 mg/L (results from at least 20 tests, spread out over as many batches, provide statistically significant data). If a lab's average is considerably less that 198, a stronger seed should be tried ... vice versa if the average is considerably more than 198. A single lab should be able to achieve a standard deviation much lower than 30.5 mg/L. A typical single lab standard deviation would be in the mid- to low- teens. A standard deviation approaching or exceeding 30.5 mg/L indicates excessive random error caused by differences in the procedure from test to test.

6. Pretreatment of Samples

a. Sample Preparation. After removing any items that are obviously not representative of the sampled water (e.g., sticks, other non-representative solids), the sample can be homogenized if necessary to make sure that a representative aliquot is used for the BOD₅ test. This can be done with a food blender on a slow speed, an aggressive stirring bar, or other device that provides thorough mixing without being overly disruptive to microorganisms in the sample as might occur if, for example, a high-speed blender is used. If BOD₅ is expected to be so high that a very small aliquot must be taken, the entire sample can be diluted such that a larger, more readily measured, aliquot may be taken. If a pipette is used to measure the sample aliquot, use of a wide-tip variety is beneficial (i.e., inside diameter of tip 1/16" to 1/8").

b. Temperature. Samples should be at 20 ± 1º C before initial DO is read. This can be done by placing sample containers in cold water in a sink if they are too warm, or in warm water if they are too cold. If only a small volume of sample is going to be diluted, it is not necessary to cool or warm the sample because, when diluted, it will not significantly affect the temperature of the dilution water which should already be 20 ± 1º C if source water is stored in the incubator.

c. Adjustment of pH. Sample pH must be in the range of 6.5 to 7.5 SU. Add sulfuric acid (H₂SO₄) or sodium hydroxide (NaOH) of sufficient concentration so that the quantity of acid or base added does not dilute the sample by more than 0.5%. For example, if the sample is one liter (1,000 mL), the acid or base should be strong enough that no more than 5 mL would be added to the sample to bring it into the range of 6.5 to 7.5 SU.

d. Dechlorination. If allowed by the discharge permit and if possible given the design of the treatment plant, BOD₅ samples taken at wastewater treatment plants using chlorine to disinfect the final effluent should be taken ahead of the chlorination point. If this is not possible, dechlorination is required, and following dechlorination, samples must be seeded (because the chlorination process kills the bacteria that otherwise would consume the BOD in the waste sample).

(1) If the sample is not highly chlorinated, dechlorination may occur naturally if samples are allowed to sit in the light for one to two hours. Samples taken from waste streams where the final effluent is dechlorinated usually do not need further dechlorination in the lab.

(2) If necessary, residual chlorine is destroyed by adding sodium sulfite (Na₂SO₃) solution. Determining how much sodium sulfite is required to dechlorinate a given amount of sample requires acidification of the sample, addition of potassium iodide, and titration with standard sodium sulfite. Since this entire procedure cannot be done on the samples that are later incubated, it must be done on a sample dedicated to that purpose.

e. Other Toxic Substances. Some wastes, particularly industrial wastes, contain metals which are toxic to the organisms responsible for oxygen depletion during the BOD₅ incubation. Such toxic materials would result in a negative bias (i.e., lower than the actual BOD₅ concentration). The presence of toxic substances can be confirmed by testing a set of serial dilutions. If the measured BOD₅ for a given sample increases significantly as the sample is increasingly diluted, a toxic substance in the sample (i.e., matrix interference) is the most likely cause. If that toxic substance cannot be avoided, its presence should be reported with results submitted by the lab. Commercial labs may know nothing about possible toxicity of samples received from most clients. In such cases, the lab might consider doing a set of serial dilutions. Although it will be too late to do anything about toxicity at the end of the five-day incubation period, the presence of a matrix interference can at least be reported to the lab client.

f. Supersaturated Samples. If initial dissolved oxygen (DO) readings with a properly calibrated DO meter (or as measured with a Winkler titration) indicate the sample contains more DO than it should for the barometric pressure and sample temperature at the time, the sample is supersaturated with DO. Supersaturation might result when the sample has been vigorously agitated just prior to the DO reading without giving air bubbles in the sample a chance to escape, or when the sample is undergoing photosynthesis. Supersaturation at the time of initial DO reading would result in a positive bias. Supersaturation can be avoided by taking the initial DO reading only on samples that are very close to 20º C, and by vigorously agitating the sample and then allowing it to settle for at least 30 minutes before taking the DO reading. A problem with supersaturation is usually indicated by high blank results (e.g., blanks exceeding DO depletion of 0.2 mg/L might be due to supersaturation), in which case the supersaturation would not be revealed by blanks exceeding a depletion of 0.2 mg/L.

g. Nitrification Inhibition. Most waste streams contain bacteria that consume nitrogen-containing organic and inorganic materials. Glutamic acid in the G/GA standard is an example of a nitrogen-containing organic material. Ammonia is an example of an inorganic material consumed by nitrifying bacteria (i.e., bacteria that convert nitrogen-containing materials to nitrite and nitrate). The materials they consume are called nitrogenous materials, or nitrogenous BOD₅. Carbonaceous BOD₅ (CBOD₅) is that part of the total BOD₅ that does not include nitrification. If CBOD₅ is to be determined rather than BOD₅, it is necessary to inhibit the nitrifying bacteria present in the sample and/or seed. This is done by adding 2-chloro-6-(trichloro methyl) pyridine (TCMP). If the lab is using the bottle method, 3 milligrams of TCMP should be added to each bottle. If the lab is using the graduated cylinder method, sufficient TCMP should be added to make the final concentration in the dilution water 10 milligrams TCMP per liter of dilution water. Pure TCMP dissolves very slowly in water. It may be advantageous to use commercially prepared reagent.

7. Procedure

a. Preparation of Dilution Water

(1) Source Water. Water used for preparing reagent solutions and dilution water, hereafter referred to as source water, must be of the highest quality. It must contain less than 0.1 mg/L copper or other heavy metals, and must be completely free of chlorine, chloramines, organic material, acids, and bases. Distilled water prepared using a glass still in the lab is sometimes suitable, although water from some stills might contain trace metals, chlorine, ammonia, or volatile organic materials making it unsuitable. Distilled water coming from a copper still is seldom suitable. Deionized (DI) water sometimes contains organic materials leached from the resin bed. Some purchased distilled waters are suitable, and some are not. If water is purchased, many labs have found steam distilled water to provide the best results. Distilled drinking water is not suitable as it generally contains chemicals (e.g., chlorine) that would adversely affect BOD₅ results. If unseeded BOD₅ blanks always run high (i.e., they always exceed 0.2 mg/L), the distilled water used to prepare dilution water should be one of the first things to investigate in trying to find the cause. Some labs have had success using tap water, others spring water ... the bottom line is, if it works (i.e., if there is no evidence of toxicity, if the average result for glucose/glutamic acid analyses is not substantially less than 198 mg/L, and results for blanks are consistently below 0.2 mg/L), use it. If it does not, try another source of water.

(2) CBOD and BOD Dilution Water. Dilution water should be prepared by different procedures depending on whether it is to be used for CBOD₅ or for BOD₅.

(a). Dilution Water for CBOD₅. Per liter of source water, add 1.0 mL each of the phosphate buffer, magnesium sulfate, calcium chloride, and ferric chloride solutions or add the contents of pre-measured packets. Aerate (i.e., saturate the solution with air) by shaking a partially filled container or by bubbling filtered air through the solution. Use a plastic or glass carboy or other container of sufficient volume that will provide sufficient water for an entire batch, but small enough that it can be easily shaken (if shaking is used to aerate). To minimize the probability that dilution water will be supersaturated with dissolved oxygen when used, store this water for at least 24 hours in the dark at 20 ± 1º C (storage in the BOD incubator is ideal). To allow free movement of oxygen to and from the container during storage, the top of the container should be loosely fitted, or replaced with a loosely-packed wad of cotton. Check stored dilution water to determine if sufficient ammonia remains after storage. Add ammonium chloride solution if necessary to bring the ammonia level to approximately 0.45 mg/L ammonia as nitrogen.

(b). Dilution Water for BOD₅. For BOD₅, aerate a container of source water as for CBOD₅ in 7a(2)(a) above, but without the buffer or nutrient solutions. It is advantageous to do this no later than the day before the BOD₅ test. Place the container in the BOD incubator to make sure the water is at 20 ± 1º C when ready for use. Approximately one hour before the BOD bottles are to be set up, add the buffer and nutrient reagents to the dilution water. Do this so as to minimize introduction of air bubbles into the dilution water container. One way of doing this is to dissolve the nutrients and buffers in a small amount of water (Hach Company's and NCL's buffer/nutrient pillows are already in solution), pour the solution down the inside wall of the dilution water container, and gently swirl (not shake, as in aerating) the container to mix the solutions. The reason for not storing BOD₅ dilution water with the nutrients/buffers already added is that a significant population of nitrifying bacteria may develop if the nutrients/buffers are present in the water. For the CBOD₅ test, the nitrifying bacteria are inhibited and therefore not a problem (see par. 4b, Standard Methods 5210B).

(3) Dilution Water Check (Blank). Regardless of how it is made, it is necessary to check dilution water with every batch of BOD₅ samples to make sure it is not causing error in the test.

(a) One check is to make sure the buffer has done its job and the pH is in the vicinity of 7.2 SU. Although not mentioned in Standard Methods 5210B, the dilution water can be checked after buffers and nutrients have been added to determine if the buffers have stabilized the pH to approximately the pH of the dilution water (6.5 to 8.5 SU). If not, the source water is suspect and alternate sources of water should be tried.

(b) Another check involves simply measuring the initial DO of a BOD bottle full of only dilution water, incubating it for 5 days, and reading the final DO. DO depletion should not exceed 0.2 mg/L. If DO depletion of the blank consistently exceeds 0.2 mg/L, suspect the source water or, maybe the dilution water was supersaturated with O₂ when the initial DO was read, indicating a need to age the water longer before setting up the BOD bottles. If DO depletion for the blank is excessive only periodically, suspect lab procedures. Contamination, calibration, equipment failure or lapse in procedure may be the cause.

b. Estimating BOD₅. A COD test can provide an estimate of BOD₅ allowing the analyst select dilution most likely to yield valid BOD₅ results. Being able to confidently predict the BOD₅ value is especially useful for commercial labs which might have no idea of the approximate BOD₅ of incoming samples.

c. Seeding.

(1) Requirement to Seed. Samples which do not already contain enough of the proper bacteria can be analyzed for BOD₅ only after addition of "seed." Seed is nothing more than a solution containing a sufficient population of suitable bacteria. Influent to a domestic wastewater treatment plant generally contains sufficient bacteria and usually does not need to be seeded. At the other end of the plant, final effluent which has been disinfected (with chlorine or ultraviolet light, for example) usually no longer has a viable bacteria population and must have bacteria added back. Other waters that usually require seeding are untreated industrial or extreme pH or high-temperature wastes. PE (performance evaluation) and standard solutions always need seeding since they contain no naturally occuring bacteria. The preferred seed is the treated, but not disinfected, effluent of that plant, the source of the samples to be tested because those bacteria are already acclimated to that environment (waste). The biochemical oxygen demand of the material in the sample will be there, the life processes of the bacteria give us a method of measuring that demand.

(2) Sources of Seed

(a) Domestic Wastewater Treatment Plants. Effluent from a domestic treatment plant also may be the best seed for samples from an industrial process or for samples not expected to have an adequate, viable bacterial population. Influent to a domestic wastewater treatment plant is generally not suitable because of the [relatively] rapidly changing character. More consistent quality is obtained from the effluent of either primary or secondary treatment processes. Some labs have found success using settled material from a clarifer. Use supernatant from the treated effluent after settling for at least one but less than 36 hours at room (20º C) temperature.

(b) Commercial Labs. Many commercial labs and some industrial discharger labs prefer to use artificial seed, such as Polybac®, Polyseed®, or BioSystems®, If these materials are used, special care must be taken in preparing the seed in accordance with the suppliers instructions. Failure to do so will result in unacceptably low results for the seed control bottle(s) (i.e., the seed control will result in a depletion of <0.6 mg/L DO per milliliter of seed) and for all samples that require seeding, including PE samples.

(c) Other Labs. Other labs might collect seed material periodically from a treatment plant and keep it viable by feeding it starch or some other nutrient while stored in the incubator. Trial and error will disclose how much nutrient is needed, and how long the seed can be maintained. Effectiveness of such stored seeds should be closely monitored by checking results from the G/GA and seed control tests and monitoring for trends indicating deterioration of the seed. Use of precision control charts is the best mechanism for monitoring performance of the seed and other aspects of the BOD₅ test.

(3) Seed Check. As already discussed, bacteria added to samples containing few viable bacteria (e.g., because of chlorination, UV treatment, extreme pH, high temperature, or toxic wastes), are called "seed" and the process is called "seeding."

(a) Check Standard. The most widely used check standard (i.e., a solution of known concentration used to check the performance of a test) for the BOD₅ test is a solution of 150 mg/L glucose (also called dextrose) and 150 mg/L glutamic acid (i.e., the G/GA test). An alternative to G/GA is a solution of 300 mg/L potassium hydrogen phthalate (KHP). An advantage of using G/GA over KHP is that G/GA is more widely recognized as the check standard for the BOD₅ test which has resulted in a large data base for test results. Also, glutamic acid contains nitrogen which provides organic material for nitrifying bacteria to consume. An advantage of the KHP test is that the 300 mg/L solution can be used as a check standard for several other tests such as COD, TOC, pH, acidity, total solids, volatile solids, and conductivity. The G/GA solution could likewise be used as checks for those tests, but the expected values are not well documented. Whichever solution is used as a standard, notice that the solution is acidic and might need to be treated with sodium hydroxide (NaOH) to bring pH into the range of 6.5-7.5 pH units.

1. Glucose/Glutamic Acid (G/GA) Check Standard

a. A standard solution of 150 mg/L of glucose and 150 mg/L glutamic acid is prepared. The standard can also be purchased ready-made.

b. Standard Methods suggests a G/GA solution should be analyzed in each batch of BOD₅ samples. This is a good idea if a lab is gathering initial data on the analysis. Once the lab is confident with its ability to do the BOD₅ test with acceptable bias and good precision, frequency of the G/GA test may be reduced to perhaps one per week, or two per month. It should be remembered, however, that any set of data is more easily defended (scientifically and legally) if a check standard and other QC samples are analyzed with each batch.

c. After at least twenty (ten is sufficient if the lab is just getting started, but 20 is preferred) G/GA solutions have been analyzed over a course of several days or weeks, the mean (average) result and standard deviation of results are calculated. Standard Methods reports that in an interlaboratory study (i.e., involving many labs), the mean result was 198 mg/L, with a standard deviation of all results of 30.5 mg/L. With that in mind, the objective established by many labs is that they be able to achieve a mean of approximately 198 mg/L, with a standard deviation of 30.5 mg/L or less. One should realize, however, that the 30.5 mg/L standard deviation was calculated from data submitted by several labs, where several analysts used several different seeds, with several variations in other test parameters. This would be expected to result in a higher standard deviation (i.e., more imprecision) than would test results from a single lab. A better goal for a single lab is to expect a mean of approximately 198 mg/L, with a standard deviation in the teens (e.g., 15 mg/L) or lower. At least one lab participating in Ecology's Environmental Laboratory Accreditation Program consistently analyzed the G/GA solution with a standard deviation of approximately 4.0 mg/L. While it is not reasonable to expect every lab to achieve such precision, it should be kept in mind that such results are possible.

2. Potassium Hydrogen Phthalate (KHP) Check Standard

a. A standard solution of 300 mg/L of KHP is prepared. This solution can be analyzed periodically as a supplement to the G/GA test, or it can replace the G/GA test. Before replacing the G/GA test, however, consideration should be given to the fact that G/GA results is most widely recognized as the primary indicator of performance for the BOD test.

b. If replacing the G/GA test, the KHP test should be run with the same frequency suggested for G/GA in paragraph 7b(3)(a)1b above.

c. Only limited data are available upon which to base quality objectives for the KHP test. In a series of 159 tests of the KHP solution over a period of approximately five years, Ecology's Manchester Lab [Washington State] achieved a mean value of 249 mg/L, with a standard deviation of 15.4 mg/L. These would be reasonable statistics to use as initial objectives until data gathered by a lab indicates other statistics might be more appropriate. If the KHP solution is used as a standard for other tests, Table 5 shows reasonable objectives for mean values for the various tests as achieved by Ecology's Manchester Lab. Precision data (e.g., standard deviations for repeated analyses) have not been determined (except for BOD₅, as indicated above).

Table 5 - Expected Values for a 300 mg/L KHP Standard

Parameter	Expected Value
BOD5	249 mg/L
COD	343 mg/L
Total Organic Carbon (TOC)	141 mg/L
Total Solids (TS)	300 mg/L
Volatile Solids (VS)	200 mg/L
Conductivity	169 mhos
Acidity	74 mg/L

(b) Seed Control Check. Another check on seed effectiveness is the seed control check. It is also used to determine what contribution the seed itself will make to the DO depletion of seeded samples. That contribution must be subtracted when calculating sample BOD₅. To do the check, set up three BOD bottles with 3, 6, and 9 milliliters of seed (or other volumes that will result in a DO depletion of at least 2 mg/L, and a retention of at least 1 mg/L after the 5-day incubation), and fill the bottles with dilution water. These are called "seed control" bottles, or sometimes "seed blank" bottles. It may be necessary to do three dilutions for the seed control until the seed material is well-characterized, at which time one bottle may give reliable results. Measure initial DO, incubate for five days, measure final DO, and calculate the DO depletion per milliliter of seed. An effective seed results in a depletion between 0.6 and 1.0 mg/L for each mL of seed. Once an effective seed has been identified and it consistently results in a DO depletion of 0.6 - 1.0 mg/L per mL of seed, it is no longer necessary to incubate three bottles with each batch and only one need be done. Table 6 shows data for an effective seed. Bottle 3 did not result in valid data because the final DO was less than 1.0 mg/L.

Table 6 - Typical Seed Control Check

	Bottle 1	Bottle 2	Bottle 3
Sample Volume (mL)
Seed Volume (mL)	3	6	9
Initial DO (mg/L)	8.9	8.9	8.9
Final (5-day) DO (mg/L)	6.2	3.3	0.8
Depletion (mg/L)	2.7	5.6	8.1
Depletion per mL of seed	0.9	<