Sewage Treatment Simulation in an Eco-Machine Prototype

P. Ganeshan, S. Greene, M. Walsh, D. Blersch, P. Kangas

Dept. Biological Resources Engineering, University of Maryland, College Park, Maryland

Waste water• Dry dog food with an analysis of 21.0% crude protein, 10.0% crudefat, 4.5% crude fiber, was used as a base to create wastewater.• Approximately 27.5g dry dog food/L mixture was allowed todecompose into a slurry in water in 121 L and 208 L covered containers.• Approximately 19 L were pumped into the trickling filter twice a day for five weeks.

Trickling Filter• Composed of two stacked 121 L containers • Upper container filled with 2” diam. bio-barrrels to facilitate microbial activity.• Lower container collects wastewater from upper filter and recirculates water back to upper container for continuous treatment.• Residence time is approximately 12 hours, after which ~19 L of treated wastewater are released to vegetation tanks. Each new addition of wastewater to the lower reservoir of the trickling filter forces an equal amount of the treated wastewater to overflow into the vegetation tanks.

Vegetation tanks• Receives 19 L of treated wastewater from trickling filter 2x a day.• Purpose is to provide additional treatment of wastewater.• Three individual tanks with water hyacinths (Eichhornia crassipes) with capacities of ~121 L, 121 L and 378 L respectively.• The storage/overflow tank has a capacity of ~100L, however volume is not constant due to evaporation.• Water circulates continuously through all tanks to provide thorough mixing within the system.

Data Collection• Five sets of water samples were collected over a five week period in each component of the living machine:

- Dog food slurry container- Lower tank of trickling filter- Third vegetation tank.

• Measurements of DO (mg/L), DO (% saturation) and Temp. (0C)were taken directly from the three locations.• Measurements of NO3

- (mg/L), NO2- (mg /L), NH4

+ (mg/L) and PO4

3- (mg/L) were taken from samples and analyzed using standard methods.• A HACH DR-2000 spectrophotometer was used to measure chemical oxygen demand (mg/L COD) from collected samples.

ABSTRACTWastewater treatment performance of an ecomachine was tested during Fall 2003. The ecomachine was housed in the Mars Society’s prototype greenhouse located on the University of Maryland campus. The purpose of the system is to serve as a test bed of life support system technology development for the society’s simulated Mars habitat in Utah. The ecomachine is a sequential treatment system composed of a recirculating trickling filter and a series of three aquatic hyacinth tanks. Total volume of the system is 210 gal and the retention time was scaled to be 21 days. The system was loaded with a simulated sewage made from a slurry of commercial dog food mixed in water. Measurements were made of temperature, dissolved oxygen, chemical oxygen demand, nitrogen compounds and phosphate. The ecomachine was effective at reducing chemical oxygen demand from more than 3500 mg/l in the simulated sewage to less than 500 mg/l in the water hyacinth tanks. In terms of nutrients, phosphate was reduced and nitrification occurred along the length of the ecomachine. These data will be utilized to improve performance of future ecomachine designs.

INTRODUCTIONThe Greenhab Project was started in 2001 by the Mars Society to explore life support technologies suitable for proposed long-term, manned space missions. The Mars Society is an organization of people interested in the exploration and settlement of the planet Mars. Human space missions require large amounts of consumables (food, water and air). A closed ecological life support system maximizes use of power from biological organisms and the sun to accomplish the recycling of human gaseous, liquid and solid waste to useful products. These concepts are based upon the ecological relationships that exist between organisms, recognizing that one organism’s waste is another organism’s food source.

A greenhouse was installed at the University of Maryland to house a living machine – a biologically-based technology suitable for multiple life support functions. The greenhouse was constructed from steel-reinforced PVC pipe (frame) and translucent polythylene sheeting (exterior) as a cylindrical module that can be combined with additional segments as required. The living machine itself is a series of bioreactor tanks, ideally composed of multiple high-diversity ecosystems, which enable the conversion and uptake of nutrients and other pollutants. Living machines have been shown to adequately and efficiently treat wastewater to acceptable levels (Todd and Josephson, 1996).

In this study, a living machine prototype was exposed to simulated sewage over a period of 5 weeks, during which time nutrient and COD dynamics were monitored in the various machine components. In addition, a simple dilution model was developed to provide a basis of comparison for the empirical data.

• The drop in phosphate concentration from sewage to trickling filter (Figure 5) can most likely be explained by incorporation of phosphate into iron oxyhydroxide compounds due to oxidation of iron as the sewage moved to the more aerobic trickling filter (Sundby et al. 1992).• The decrease in dissolved oxygen (Figure 6) from trickling filter to vegetation tanks is probably too small to explain the coincident increase in dissolved phosphorus (P).• The increase in P concentration from trickling filter to vegetation tanks could be explained by changes in pH, as a small increase in pH can create large increases in dissolved P due to desorption from particles (Stumm and Morgan 1970). Future monitoring efforts should include pH • To effectively remove P from wastewater, it must be ensured that P remains bound in sediments, which requires relatively constant redox conditions and pH.

• Nitrate became an increasingly important component of TIN in the trickling filter and vegetation tanks.• This may be evidence of nitrification

• There appeared to be a decline in TIN over time throughout the living machine.• This may be due to decreased remineralization of organic nitrogen in “sewage” as microbial processes slowed due to decreasing temperatures• An analysis of organic nitrogen in the living machine would provide more insight into nitrogen dynamics and permit an analysis of total nitrogen removal by the living machine.• Measurements of denitrification would also add to the analysis.

Figure 7: Total inorganic nitrogen concentration in living machine components.

CONCLUSIONS

• The living machine significantly decreased the COD of sewage inputs, as compared to zero-treatment model outputs.• The living machine has the potential to effectively removephosphate, but optimum pH and redox conditions must beMaintained.• There was strong evidence that nitrification occurred, and perhaps evidence of nitrogen removal.

• Phosphate exhibited a consistent spatial pattern.• Highest concentrations observed in “sewage”• Lowest concentrations observed in trickling filter• Intermediate concentrations observed in vegetation tanks

Figure 4: Phosphate concentrations in living machine input sewage, trickling filter and vegetation tanks.

Figure 5: Example of spatial patterns in dissolved phosphate concentration in living machine components.

RESULTS AND DISCUSSION: COD

Figure 2: Results of simple dilution model, simulated with STELLA, showing the expected COD concentration over time in the living machine assuming absolutely no COD removal by the living machine.

Figure 3:COD levels in input sewage, trickling filter and vegetation tanks over time.

• The STELLA simulation graph (“no treatment” scenario, Figure 2) shows how the 3800 mg/L input COD increases over time in the vegetation tanks and trickling filter. • Actual data (Figure 3) shows substantial COD removal with time in both the trickling filter and vegetation tanks, especially compared to the “no-treatment” scenario of Figure 2.• COD levels begin to rise as the system is overloaded with sewage (Figure 3).

REFERENCES

Stumm, W. and J.J. Morgan. 1970. Aquatic chemistry: an introduction emphasizing chemical equilibria in natural waters. Wiley Interscience, New York.

Sundby, B., C. Gobeil, N. Silverberg, and A. Mucci. 1992. The phosphorus cycle in coastal marine sediments. Limnology and Oceanography. 376(6):1129-1145.

Todd, J. and B. Josephson. 1996. The design of living technologies for waste treatment. Ecological Engineering. 6: 109-136.

ACKNOWLEDGEMENTS

• The Nutrient Analytical Services Lab (NASL) at the Chesapeake Biological Laboratory, Solomons, Maryland.

•The Mars Society, especially Gary Fisher of the Philadelphia Chapter.

Greenhab greenhouse segmentat the University of Maryland.

Vegetation tanks and trickling filter inside the Greenhab greenhouse.

A student collects data from vegetation tanks inside the Greenhab greenhouse.

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

0 10 20 30 40 50Time (days)

CO

D (

mg

/L)

Input Conc

Trickling Filter

Aquatic Tanks

0

500

1000

1500

2000

2500

3000

3500

4000

4500

0 5 10 15 20 25 30 35

Time (days)

CO

D m

g/l

Vegetative Tanks

Trickling Filter

Input Sewage

Contribution to "sewage" TIN by various N species

0%10%20%30%40%50%60%70%80%90%

100%

8-Oct 15-Oct 22-Oct 29-Oct 5-Nov

Sample Date

Per

cen

t o

f T

IN

NH4

NO2

NO3

Contribution to trickling filter TIN by various N species

0%10%

20%30%40%

50%60%70%80%

90%100%

8-Oct 15-Oct 22-Oct 29-Oct 5-NovSample Date

Per

cen

t o

f T

IN

NH4

NO2

NO3

Contribution to vegetation tank TIN by various N species

0%10%20%30%40%50%60%70%80%90%

100%

8-Oct 15-Oct 22-Oct 29-Oct 5-Nov

Sample Date

Per

cen

t o

f T

IN

NH4

NO2

NO3

Figure 8: Percent composition of Total Inorganic Nitrogen in living machine components.

0

50

100

150

200

250

300

8-Oct 15-Oct 22-Oct 29-Oct 5-Nov

Date

[PO

43

-] (

mg

/L)

Dogfood Slurry

Trickling Filter

Vegetation Tanks

0

5

10

15

20

25

30

35

40

45

50

8-Oct 15-Oct 22-Oct 29-Oct 5-Nov

Date

[TN

] (m

g/L

)

Dogfood Slurry

Trickling Filter

Vegetation Tanks

Trickling Filter

Vegetation Tanks

Dogfood Slurry

0

1

2

3

4

5

6

7

8

9

10

Component

Dis

so

lve

d O

xy

ge

n (

mg

/L)

Dogfood Slurry

Trickling Filter

Vegetation Tanks

0

50

100

150

200

250

Component

Ph

os

ph

ate

(m

g/L

)

Figure 6: Example of spatial patterns in dissolved oxygen concentration in living machine components.

MATERIALS & METHODSThe Mars Greenhab living machine consists of three basic parts and systems (wastewater, trickling filter and vegetation tanks) that are combined to make a complete unit as shown in Figure 1. The purpose of the Greenhab living machine is to improve water quality.

Phosphates

Nitrogen (cont.)

Nitrogen