Carbon Source Comparison Study

by Justin Hall of UW-Steven’s Point.

Comparison of three food supplements and their effectiveness at wastewater treatment in batch reactors. Food supplements tested include; Molasses, Old Roy dog food, and Aquafix’s SmartBOD.

Reactor Design

Three batch reactors were constructed using 6” PVC pipe with a total volume of approximately 7L and a working volume of 5L.  The reactors each had independent pH and dissolved oxygen controls, in addition to monitoring temperature.  A custom microcontroller interface was designed to control the reactor set points, and to view real-time sensor output.

Wastewater was obtained from the Stevens Point, Wisconsin municipal wastewater treatment facility.  The wastewater was diluted to 1500 mg/L MLSS concentration using city tap water prior to the study start.

Figure 1: Design schematic for batch reactors

Experimental Conditions

Three different food supplements (carbon sources) were used; molasses, Old Roy dog food, and SmartBOD.  The reactors had the following controls; pH, dissolved oxygen, and mixing.  Temperature was monitored but not controlled.  Mixing was achieved from an overhead mixer, and aeration (when active).  Conditions for the reactors are listed in Table 1.

Table 1: Experimental conditions

Parameter/Condition Setpoint
BOD of influent To achieve 250 mg BOD / L * day-1
pH 7.30-7.50
DO 2.8 mg/L DO
Temperature Not controlled
Feed / Waste cycles per day 3
Volume decanted per day 1L
MLSS wasted per day Varied per reactor, held MLSS at 1500 mg/L
A 5% solution of sodium carbonate and 0.1M hydrochloric acid were used to control pH, and an aquarium pump was used to control dissolved oxygen. Fresh solutions were made as necessary.

The amount of each food supplement differed based on supplied BOD values. Table 2 illustrates volumes added to each reactor per day to achieve a BOD concentration of 250 mg BOD / day.

Table 2: Volumes of food added to reactors

Food Fource BOD Food Added Food Slurry Dilution Water
mg/g g/day mL/day mL/day
Molasses 486 2.57 100 900
Old Roy 43 29.1 1000 0
SmartBOD 427 2.93 100 900

Concentrated mixtures of the molasses and SmartBOD were made so that the required dosage rate in grams were added in 100mL. The Old Roy dog food had to be made up in a larger volume due to the higher viscosity of the slurry. The mixtures of the SmartBOD and dog food were continually mixed with an overhead mixer, and were stored at 4°C. Food mixtures were made fresh at least every other day.

Each reactor completed three cycles daily consisting of; of treatment, settling, decant, and feed/dilution phases. Each cycle lasted eight hours. Reactors settled for 30 minutes before decanting occurred. The feed and dilution phases occurred simultaneously.

Test Methods

Reactors were tested for the following; ammonia, reactive phosphorus, chemical oxygen demand, flora count, mixed liquor suspend solids, and sludge volume index. Table 3 lists the method used for each test.

Table 3: List of methods

Test Method
Ammonium (NH3-N) SM4500 NH3H
Reactive Phosphorus (PO43-) SM4500 PG
Chemical Oxygen Demand (COD) SM 5220 D
Flora count Toni Glymph
Mixed Liquor Suspended Solids (MLSS) SM2540 D
Sludge Volume Index (SVI) SM2710 D
Spectroscopy In House

Ammonium and Reactive Phosphorus

Samples were collected from reactor effluent, filtered through a 0.45 µm membrane filter and preserved to a pH of < 2.0 with 1:1 H2SO4. Samples were analyzed by the Water and Environmental Analysis Laboratory located on the University of Wisconsin – Stevens Point campus.

Chemical Oxygen Demand

Reactor effluent samples were collected, filtered, and preserved with H2SO4 and refrigerated at 4°C until the time they could be analyzed. Chemical oxygen demand samples were analyzed in house.

Flora Count

A flora count was completed, using a method developed by Toni Glymph. Three slides were analyzed for each reactor’s MLSS. The different species of flora observed were tallied and averaged over the three slides. These results are reported as percent of total flora.

Mixed Liquor Suspended Solids

Samples of Mixed Liquor were removed from the reactor. The volume removed from the reactor was replaced with tap water to keep the 5L volume of the reactors consistent. Reactors were wasted as needed to keep MLSS at 1500 mg/L.

Sludge Volume Index

1L of the mixed liquor was removed from the reactor and allowed to settle for 30 minutes. The samples were then returned to the reactors.


An attempt was made to compare the MLSS values obtained from SM2540 D to uv-vis spectroscopy. Data was collected during the study, however it proved to be inconclusive. It was realized that during the study a background correction needed to be obtained to correct for color that developed in the reactors. Work on the comparison of absorbance values to MLSS concentrations will continue in future studies while correcting for background color.

Results & Discussion

Ammonium and Reactive Phosphorus

Effluent ammonium and reactive phosphorus were analyzed. Table 4 shows concentrations of each sample analyzed. It is unclear why the initial samples had high levels of ammonium and phosphorus in reactor effluent compared to later sample dates. It is possible that there was still carry over from the MLSS seed obtained from the Stevens Point wastewater treatment plant.

Ammonium did appear to accumulate in the reactor containing the Old Roy dog food, with low levels observed in the rest of the samples. Phosphorus accumulated in the reactor containing the dog food and molasses. Besides the initial sample, low concentrations of ammonium and phosphorus were observed with SmartBOD.

Table 4: Ammonium and reactive phosphorus concentrations in reactor effluent

NH3-N (mg/L)

Date Molasses Old Roy SmartBOD
11/10/2016 0.07 0.39 <0.01
11/15/2016 0.02 0.02 <0.01
11/21/2016 0.05 0.43 0.04
11/28/2016 0.04 0.99 0.01

PO43- (mg/L)

Date Molasses Old Roy SmartBOD
11/10/2016 1.82 1.86 0.383
11/15/2016 0.144 0.112 0.032
11/21/2016 0.195 0.155 0.039
11/28/2016 0.143 0.168 0.031
Figure 2: NH3-N concentrations in reactor effluent
Figure 2: PO43- concentrations in reactor effluent

Chemical Oxygen Demand

Effluent chemical oxygen demand was tested as a metric for the efficiency of treatment of the wastewater in the reactors. The reactor with the Old Roy dog food showed a clear increase in COD throughout the study, to nearly 1000 mg/L. This would indicate that treatment was not complete during reactor cycles and continued to accumulate. Both the molasses and the SmartBOD had low levels of COD in their effluent. All of the SmartBOD values, except for November 3rd and 9th were below 100mg/L COD, the rest were near or below 50 mg/L COD. This would indicate continuous, effective removal of the COD from the wastewater. Table 5 and Figure 4 contain COD results.

Table 5: COD concentrations in reactor effluent
Date Molasses Old Roy SmartBOD
11/3/2016 156.7 204.7 118.0
11/9/2016 155.4 468.1 115.8
11/12/2016 131.0 764.8 52.0
11/15/2016 125.2 832.9 39.0
11/22/2016 140.4 933.8 50.6
11/28/2016 200.9 909.4 46.2
Figure 4: Chemical oxygen demand in reactor effluent

Flora Count

Microscope analysis of the MLSS was completed to look at the flora present in each reactor. The types of flora present can indicate the relative age, and health of sludge. Healthier, younger sludges tend to have more flagellates and ciliates, where older, less healthy sludges tend to have more amoebas, and metazoa.

Table 6: Flora count of MLSS. Results listed as a percent of total count. Molasses (Rx1), Old Roy (Rx2), and SmartBOD (Rx3)


Date Rx1 Rx2 Rx3
11/4 5.7% 0.0% 0.0%
11/15 0.0% 4.5% 0.0%
11/21 3.2% 4.1% 0.0%
11/28 2.3% 0.0% 0.0%


Date Rx1 Rx2 Rx3
11/4 9.4% 9.3% 5.5%
11/15 14.6% 20.5% 21.2%
11/21 16.1% 12.2% 12.9%
11/28 13.6% 8.9% 11.5%

Free Swimming Ciliates

Date Rx1 Rx2 Rx3
11/4 9.4% 2.3% 16.4%
11/15 12.5% 2.3% 7.1%
11/21 16.1% 2.0% 7.1%
11/28 13.6% 3.6% 12.8%

Crawling Ciliates

Date Rx1 Rx2 Rx3
11/4 0.0% 0.0% 1.8%
11/15 2.1% 0.0% 2.4%
11/21 0.0% 0.0% 0.0%
11/28 0.0% 0.0% 0.0%

Stalked Ciliates

Date Rx1 Rx2 Rx3
11/4 5.7% 30.2% 9.1%
11/15 12.5% 18.2% 28.2%
11/21 12.9% 12.2% 28.6%
11/28 11.4% 19.6% 35.9%


Date Rx1 Rx2 Rx3
11/4 69.8% 58.1% 67.3%
11/15 58.3% 54.5% 41.2%
11/21 51.6% 69.4% 51.4%
11/28 59.1% 67.9% 39.7%

Sludge Volume Index

Sludge volume index (SVI) is an additional metric that can be monitored to characterize the health of wastewater sludge.  SVI is the volume of solids that are occupied after a 30-minute settling period, in milliliters.  Table 7 has general guidelines for the results from a SVI test. 

Table 7: General guidelines for SVI

SVI Sludge Characteristic
< 80 ML/g Dense with rapid settling. Old and potentially over oxidized sludge
100 to 200 mL/g Good quality effluent. Medium settling, irregular flocs
> 250mL/g Slow settling, poor floc formation

The reactor fed with molasses for the first part of the study had a slightly higher SVI then SmartBOD. During the test, good settling was noted with both reactors. However, after a slow increase in SVI, the SVI rapidly decreased to less than 100 mL/g. A dense mass was observed after the 30-minute settling period. During this same time, it was noted that during the MLSS testing for the reactor fed with molasses became difficult to filter. A possible cause for this is from the formation of filamentous bacteria. SmartBOD maintained a SVI in the ideal range of 100 to 200 mL/g for the duration of the study.

Old Roy dog food saw a continual increase in stability. Poor floc formation and poor settling were observed during testing. Table 8 and Figure 5 illustrate the SVI from the study.

Table 8: Sludge volume indices

SVI (mL/g)

Date Molasses Old Roy SmartBOD
11/4/2016 217 363 157
11/7/2016 244 341 140
11/11/2016 230 397 157
11/16/2016 267 360 163
11/18/2016 263 457 170
11/21/2016 250 513 168
11/24/2016 150 553 137
11/28/2016 93 528 167
11/30/2016 70 619 158

Figure 5: Sludge volume indices


The addition of SmartBOD as a food supplement (carbon source) outperformed molasses and Old Roy dog food when comparing the metrics of COD removal, nutrient accumulation and settling properties. The clearest advantage is when comparing it to the use of Old Roy dog food. The dog food reactor had high levels of solids, nutrient build up, and high levels of COD in the reactor effluent. All of these factors indicate that there was only partial treatment in the reactor containing the Old Roy dog food.

SmartBOD and molasses have similar levels of nutrient removal and a slightly less COD removal efficiency. The SmartBOD outperforms the molasses when it comes to the settling of the sludge. The SVI for the SmartBOD is right in the ‘sweet spot’ indicative of healthy sludge.

Concluding, the evidence from this study supports that the SmartBOD is a more efficient carbon source when compared to Old Roy dog food and molasses.


Subscribe To Our Water Solutions Newsletter

Join our mailing list to receive the latest news and updates about water treatment solutions. We'll never spam you and your private email information is safe with us.

You have Successfully Subscribed!