Progress Report

CICEET Progress Report for the period 3/16/06 Through 9/15/06

Project Title: Wastewater Treatment to Minimize Nitrogen Delivery from Dairy Farms to Receiving Waters
Principal Investigator(s): Nancy G. Love, Ph.D.
Additional Investigator(s): Katharine F. Knowlton, Ph.D., Barth F. Smets, Ph.D.
Project Start Date: November 1, 2004 (Delayed due to date of funding arrival).

Figures


Figure 1

Figure 1


Figure 2

Figure 2


Figure 3

Figure 3


Figure 4

Figure 4


Figure 5

Figure 5


Figure 6

Figure 6


Figure 7

Figure 7


Figure 8

Figure 8


Figure 9

Figure 9


Tables


Table 1

Table 1


Table 2

Table 2


Project Objectives for This Reporting Period
Objectives
The current objectives are listed below. Those objectives that pertain to this reporting period are highlighted in RED.

Objective 1. Demonstrate the ability of a fermentation-fed nitritation/denitritation (NIT/DENIT) process to remove nitrogen from anaerobically stabilized dairy wastewater.

Objective 2. Experimentally determine the operating parameters and treatment capacity for aNIT/DENIT-based treatment system.

Objective 3. Characterize the liquid, solid and gaseous effluent from the NIT/DENIT process.

Objective 4. Develop practical design equations and performance curves for the NIT/DENIT-based treatment system.

Objective 5. Ensure environmental compatibility by calculating a nitrogen balance for a model concentrated dairy operation located in the NERR region.

Tasks to meet objectives
The following tasks are required to meet the objectives that were addressed during this reporting period.
1. Construction and operation of a pilot-scale fermentation system to produce volatile fatty acids, based on outputs from the laboratory-scale experiments.

2. Construction and operation of a pilot-scale nitritation-denitrification (NIT/DENIT) sequencing bioreactor to achieve nitrogen removal. This reactor is fueled by the output from the fermentation reactor.

3. Visit and evaluate nitrogen cycling on multiple dairy farms, including the model farm in Winyah Bay. These balances will serve as models in which we will later evaluate the impact of implementing nitrogen removal on a dairy farm.

Progress on Tasks
The pilot plant system has been constructed (Objective 1), along with the shed that houses the system (our previous barn blew down April 2006 in a wind storm; the new shed is more robust and was built to last longer for future dairy waste treatment studies). Figure 1 shows the pilot-scale system housed in the shed. The fermentation reactor is operating and generating organic acids (results presented in Section 3 below) while we have encountered start up-challenges with the nitrogen removal reactor (results presented in Section 3 below and challenges in section 1.4 below). To assist with identifying our operation strategy, a modeling exercise was performed as a starting point for objective 2. The output from this modeling exercise forced us to revise our experimental reactor configuration and ended up being a critical tool in our study. We have been monitoring both reactors in the pilot-scale system (Objective 3) and will report on the status of that monitoring effort in Section 3. Finally, three collaborator farms within the Winyah Bay watershed were visited, and data were collected to allow construction of Whole Farm Nutrient Balances (Objective 5). Specifically, data were collected on nitrogen imports to the farm (primarily in the form of feed and fertilizer) and nitrogen exports from the farm (primarily in the form of milk). All three farms show significant nitrogen imbalances, with nitrogen imports exceeding exports. Final data is being collected on the purchase and sale of livestock from the farms (a relatively small contributor). We will report these results in the next reporting period.

Difficulties
Our primary difficulty with the manure treatment system has been to get the low dissolved oxygen levels in the NIT/DENIT reactor to stabilize sufficiently so that we can achieve nitritation. We also encountered a cold spell in Blacksburg that corresponded with an electrical outage, which resulted in a long term cold stress event on the reactor system in mid-October. This seems to have disabled the nitrification biomass substantially and we reseeded the system in late October. Furthermore, we have purchased one of Hach’s new combined luminescent dissolved oxygen (DO) and pH monitoring and control devices to provide 24 hour control of DO and pH, which is mandatory for achieving stable nitritation.

Data Generated to date

Project Objectives for Next Reporting Period

Objectives
Objectives 1, 3 and 5 listed under section 1.1 will be our primary focus during the next reporting period.

Tasks to Meet Objectives
We hope to complete objective 1 (successfully demonstrate nitrogen removal with the system) by the next reporting period. Installing the DO and pH control unit will be critically helpful in this regard. We have a reliable space heater system operating in the shed now to maintain viable temperatures over the wintertime. Objective 3 will be ongoing through the next reporting phase, including resolving methods to analyze the offgas. The initial mass balance calculation for our sample farm in Winyah Bay has been completed (objective 5). Balances for all three collaborator farms will be fine tuned by the next reporting period and will be presented at that time. If time allows, we will initiate the development of the simplified design equations identified in objective 4.

Work plan to Meet Objectives
The Dairy Waste Management Research Team meets biweekly to discuss progress on this and other related Dairy Waste projects (including follow up work to our other CICEET-funded project related to P removal in dairy waste). In addition, Dr. Love meets with Mr. Jason Beck, the graduate student assigned to the pilot plant project. We have hired an undergraduate assistant to help Mr. Beck, which is freeing up his time to process samples and to analyze data. We will be primarily focused on monitoring and assessing the nitrogen removal pilot plant.

Anticipated Success in Meeting Project Objectives
The success of these project objectives will produce a system with lower oxygen demand and sludge production. Compared with conventional nitrification, air requirements for nitritation are 25% less while carbon requirements for denitritation are approximately 40% lower than that for denitrification from nitrate (Lai et al., 2004). The carbon requirements were calculated and are reported below in the “Preliminary Data” section. These calculations demonstrated that a configuration change was needed in the design mentioned in the previous report (Figure 2). The rationale for this change is discussed below in Section 3. The preliminary modeling work has provided us with confidence in the success of the newly proposed system. Furthermore, we are confident in the nitrogen removal capabilities of these new systems because they utilize conventional microorganisms that have been evaluated, are known to be reliable, and have been grown in our laboratory in the past. This coupled with the critical DO and pH control device will allow us to achieve stable N removal in the next reporting period. The initial calculation of whole farm nutrient balance for the case study farms (complete except for data on purchase and sale of livestock, a small portion of the balance) indicates significant excesses of N on case study farms. More N is being imported, primarily in the form of feed, than is being exported in product (primarily milk). This is similar to observations for dairy farms in Utah and Idaho (Spears et al., 2003) in which only 35% of imported N was exported in product. Other authors have reported exports as low as 15% of imports (Aarts et al., 1992) and up to 46% of imports (Bacon et al., 1990). Nitrogen imported on to the farm and not captured in meat and milk will be lost through leaching, runoff, or volatilization, as N does not accumulate in soil to any great extent. The observed N imbalance observed on our collaborator farms indicates a clear need for technologies to control ammonia loss from farms in our target watershed.

Overall Project Timeline Update
We requested and received a one year no cost extension for this project. We anticipate operating the pilot plant system until July, at which point we will complete our analysis and final report for the project. Our general timeline is:

December 31, 2006 Achieve stable nitrogen removal
January 31, 2007 Evaluate extent of ammonia oxidation versus nitrite oxidation in reactor. Complete first version of mass balance on Winyah Bay Dairy Farm.
March 31, 2007 Evaluate performance at loading #1
April 30, 2007 Evaluate performance at loading #2
May 31, 2007 Evaluate performance at loading #3. Develop first version of design equations. Revise mass balance on Winyah Bay Dairy Farm.
July 31, 2007 Repeat at least two of the loadings and refine design equation development
September 1, 2007 Complete final report for submission to CICEET

preliminary Data
3.1 Fermentation
The stoichiometry of the fermentative process involves using organics in the dairy manure as both the electron acceptor and electron donor along with seeded bacteria to produce volatile fatty acids (VFAs). The VFAs will then be used as the electron donor in the denitritation process.

The lab fermenter was constructed as a sequencing batch reactor (Figure 1) with a solids and hydraulic retention time (SRT and HRT) of 2 days. The total volume of the laboratory-scale fermenter was 8 L. The fermenter was started on January 26, 2006 and had the following constituents as the starting materials:

  • 2 L of anaerobic digester sludge from Christiansburg, VA WWTP
  • 2 L return activated sludge (RAS) from Christiansburg, VA WWTP
  • 4 L solids-separated and diluted dairy waste

Operation of the lab scale fermenter ended in early June 2006. The VFA production averaged 1,850 + 200 mg/L of acetic acid as COD (Figure 3). This reported concentration is based on a COD of 1.08 g COD/g acetic acid (Grady et al., 1999) The fermenter also produced 700 mg/L NH3-N. This ends up being a very important element that incited a change in our experimental reactor configuration. Finally, as reported previously, the TSS and VSS values in the fermenter were relatively constant at 9,500 and 8,000 mg/L, respectively.

The pilot scale fermenter was set up the same as the laboratory fermenter. This system has been running for two months at this point. Data are not available due to a problem with the gas chromatograph used to analyze these samples. They will be presented in the next report.

3.2 Aerobic Nitritation
The stoichiometry of the nitritation process is:
     AOB (Ammonia Oxidizing Bacteria) NH4+ + 1.5O2 --> NO2- + H2O + 2H+ [eq 3.2a]
     NOB (Nitrite Oxidizing Bacteria) NO2- + 0.5O2 --> NO3- [eq 3.2b].

The two-step nitrification process converts ammonia to nitrite, then nitrite to nitrate by two distinct genera of bacteria. Conventionally, the effluent is then sent to an anoxic denitrification step where nitrate is used as the electron acceptor and is converted to nitrogen gas (N2). The goal of the NIT/DENIT process is to skip the oxidation of nitrite to nitrate, thereby saving energy. As stated previously, achieving nitritation by selecting for ammonia oxidizing bacteria (AOBs) while inhibiting nitrite oxidizers to nitrate (NOBs) has several advantages. These include a lower oxygen demand, less demand for an exogenous electron donor to fuel denitrification, and a slightly lower volume of sludge production. Since nitritation is an acidifying process, pH control is very important and must be kept above 7.0. For our laboratory system, the pH was controlled using a pH controller that activated a pump and added a base (sodium bicarbonate) when acidification began in order to keep the pH relatively constant. Similarly, we have the ability to add base in the pilot-plant system as needed; however, since denitrification forms alkalinity and dairy waste is high in alkalinity, the base addition will be minimized.

A low dissolved oxygen concentration in the nitritation reactor was maintained to inhibit nitrite oxidizing bacteria. A low DO level was achieved by using a mixture of air and nitrogen at a constant and well defined flowrate. Frequent DO measurements were taken to ensure that the target DO was being achieved. The reactor was monitored for ammonia, nitrite, nitrate and pH.

To date, we have had minimal nitrite formation or nitrogen removal from the pilot plant system. Upon troubleshooting recently, we considered several possible reasons:
1. The solids retention time is too short. Even though the simulation discussed next (section 3.4) indicated an optimal retention time to achieve N removal, that model uses kinetic parameters typical of domestic wastewater treatment systems. If we are dealing with more slowly degradable material, some of these kinetic parameters could be off, thereby causing nitrifier washout. Solution: increase solids retention time.

2. The DO level is insufficient to support any nitrification and needs to be higher. We are feeding a fairly high COD load to the reactor so it is conceivable that we need to poise the NIT/DENIT reactor at a slightly higher DO than we have been. Solution: increase DO and implement DO control and monitoring system.

3. Inhibitors to nitrification exist in the dairy waste. Perhaps there are byproducts of the dairy farming process that are present in the manure and inhibit nitrification. We consider this scenario unlikely, especially given that our previous phosphorus removal CICEET project showed that nitrification occurred readily with dairy waste and had to be inhibited to prevent it. Solution: to test for this, we can perform a series of respirometry tests. If the respiration rate increases as dairy waste is more diluted, it is a direct indication of inhibition.

These scenarios were considered at a recently biweekly team meeting, and we elected to address all three. First, we are implementing both DO and pH control with our new Hach metered system. Second, we are reseeding the NIT/DENIT reactor and increasing the overall retention time from 6 days to 12 days. Third, we are in the process of running a series of respirometric analyses of the dairy waste to see if, in fact, an inhibitor has been added to the waste since our previous CICEET dairy waste project.

3.3 Previous design modification and new proposed pilot scale design
The NIT/DENIT system originally proposed to CICEET included three separate reactors. The first reactor was a fermenter operated in a similar manner to how the lab fermenter was operated. Second, we planned to operate a nitritation reactor fed effluent from an anaerobically stabilized lagoon so that it converted NH3 to NO2-. Finally a separate denitritation reactor would receive the effluents from both the nitritation reactor and the fermenter where heterotrophic bacteria would remove nitrogen via heterotrophic denitritation (Figure 4). This configuration was assessed using a modeling software package called Biowin®. The original objective of the modeling exercise was to determine the optimum flowrate of the fermenter needed to achieve a minimum nitrogen level in the denitritation effluent. However, the results showed that the addition of the fermenter effluent actually increased the total nitrogen in the final effluent at all values of fermenter flow (Figure 5). This is due to the fact that the fermenter effluent contains NH3 (approximately 700 mg/L as N) and use of the configuration highlighted in Figure 4 would essentially add bioavailable nitrogen to the denitritation reactor along with VFAs. The presence of this bioavailable nitrogen in the fermenter effluent makes the previous design inadequate.

The substrates of NH3 and VFAs need to be utilized by autotrophic AOBs and heterotrophic denitrifiers in the same reactor because the two substrates are contained within the same fermenter effluent. Therefore, a new system configuration was designed that combines the nitritation and denitritation reactions from the previous design into a single reactor (NIT/DENIT). This is achieved by operating the reactor with cyclic aeration (Figure 6). With this configuration, ammonium will be converted to nitrite during the aerobic phase with a target DO range of 0.7 to 1.4 mg/L (Ruiz et al., 2003). The fermenter effluent will be pumped in during anoxic phases (no DO) and the VFAs present will be used as a carbon source for heterotrophic denitritation. By cycling between microaerobic and anoxic conditions, we expect to achieve nitrogen removal of both the ammonia in the original dairy waste as well as that generated by the fermenter.

Hand calculations were completed to confirm the adequacy of this design, followed by a Biowin® simulation. An average concentration of 1,850 + 200 mg/L of acetic acid as COD and 700 mg/L of NH3-N was used to reflect the fermenter effluent. Using stoichiometry, the required ratios of readily biodegradable COD (rCOD) as VFA:NH3 for complete heterotrophic N removal was found to be 2 if the electron acceptor is NO2- and 4.8 if the electron acceptor is NO3- (Grady et al., 1999) (Table 1). The measured rCOD as VFA:NH3 ratio in the fermenter effluent was found to be 2.6. Therefore, N removal is possible with the newly proposed design if nitritation is achieved.

3.4 System performance and operating conditions for revised system configuration
Solids retention time (SRT), flow rate, reactor volumes and percent aeration (only applicable to the NIT/DENIT reactor) were considered when designing this system. First, fermenter SRT and effluent properties were determined from the lab fermenter experiment. Next, the NIT/DENIT reactor volume and overall scale of the system were determined based on equipment boundaries and feasible operating procedures. Finally, the percent aeration, SRT of the NIT/DENIT reactor and the system flow rate was determined through the use of Biowin® modeling.

Total SRTs of 2, 3, 4, 6 and 8 days were chosen for testing in the NIT/DENIT reactor. Several modeling simulations were performed for each SRT with the goal of testing the effect of percent aeration on effluent nitrogen levels (Figure 7). The results show that overall minimum nitrogen concentrations decrease as the total NIT/DENIT SRT increases. For example, at an SRT of 6 days, Figure 8 shows that a percent aeration of 30 percent was appropriate to achieve the minimum level of nitrogen, based on assumed parameters used in the simulation. An SRT of 8 days actually produced a slightly lower level of effluent nitrogen. We initially operated the system at a total SRT of 6 days to determine if we could get the simulated performance plus we are concerned about producing hydrogen sulfide gas if we increase the SRT too much. Information on this experimental setup are provided in Table 2.

The performance of the system at a total SRT of 6 days has been unsatisfactory. As shown in Figure 9, we have been unable to sustain ammonia oxidation (nitritation, our goal) or even nitrite oxidation, despite maintaining a DO of around 0.5 mg/L. This is curious, as we have no trouble achieving nitritation in laboratory systems at this level, or in previously maintained phosphorus removal treatment systems functioning on dairy waste (previous CICEET project). As noted in section 3.2, item 1 above, we are incorporating DO control and monitoring with our new Hach monitoring unit, although we believe our system maintained relatively constant DO during the previous phase. We will also be able to retain constant pH, which we may not have achieved previously. Furthermore, we are increasing the total system SRT. If these changes result in the desired level of nitritation, we can assume that we were washing the ammonia oxidizing bacteria out of the reactor at a total SRT of 6 days. This suggests that key kinetic parameters for our BioWin® simulations will have to be estimated in order to improve the models ability to predict performance.

Dissemination
Section 4 - Dissemination

Publications: none yet

Workshops: none

Conferences:
W. Khunjar, P. Sweetman, N. G. Love, K. F. Knowlton, B. F. Smets. Treatment of Anaerobically Stabilized Dairy Waste with an Oxygen Limited Autotrophic Nitrification plus Denitrification (OLAND) Fixed Film Reactor: Startup and Maintenance Issues. Virginia VWEA/AWWA Water Jam Conference, September 29, 2005.

Beck, J., Love, N. G., Knowlton, K. F., Ogejo, J., Gungor, K. Nitrogen removal from dairy waste using deammonification fueled by fermented dairy manure, accepted for poster presentation at the American Society of Agricultural and Biological Engineers, June 17-20, 2007, Minneapolis, MN.

Manuals, Protocols none
Outreach Activities: none
Contact with End Users: The PIs have communicated with the primary end-user, Dale Gardner of the Virginia State Dairyman’s Association regarding our progress. Furthermore, Dr. Knowlton has performed assessments with the primary dairy farm in Winyah Bay which is being used to perform the nitrogen.

Patent, Copyright, Invention Disclosure Activity: None at this time. We have contacted our university intellectual property officer and a review is underway to determine if IP exists.

Expenditures
Based on a 3 year project (no cost extension considered), the project is 75% complete based on time and 86% complete based on expenditures. Funding to pay the primary graduate student on this project will be used up by April 2007, after which we will use internal extension funding to support him until the new end date of August 2007. This will ensure that the project is completed. His thesis will serve as the basis for the final report.

End User Advisor Feedback
Name: Dale Gardner, Executive Secretary
Organization: Virginia State Dairyman’s Association
Location: Harrisonburg, VA
Phone Number: 540-434-6722
Email: vamilk4u@gte.net

A letter has been requested.

References
Aarts, F. M., E. E. Biewinga, and H. V. Keulen. 1992. Dairy farming systems based on efficient nutrient management. Netherlands J. Agric. Sci. 40:285-299.

Bacon, S. C., L. E. Lanyon, and J. R. M. Schlauder. 1990. Plant nutrient flow in the managed pathways of an intensive dairy farm. Agron. J. 82(July-August):755-761.

Grady, J. C. P. L., G.T. Daigger and H.C. Lim. (1999). Biological Wastewater Treatment, Marcel Dekker, New York, NY.

Lai, E., S. Senkpiel, D. Solley, J. Keller. 2004. Nitrogen removal of high strength wastewater via nitritation/denitritation using a sequencing batch reactor. Water Science and Technology. 50(10): 27-33.

Ruiz, G., D. Jeison, and R. Chamy. 2003. Nitrification with high nitrite accumulation for the treatment of wastewater with high ammonia concentration. Water Research. 37:1371-1377

Spears, R. A., R. A. Kohn, and A. J. Young. 2003. Whole-farm nitrogen balance on western dairy farms. J. Dairy Sci. 86:4178-4186.

Strous M, J. J. Heijnen, J. G. Kuenen, and M. S. M. Jetten. 1998. The sequencing batch reactor as a powerful tool for the study of slowly growing anaerobic ammonium-oxidizing microorganisms. Appl Microbiol Biotechnol. 50:589­596.

Van de Graaf, A.A., P. de Bruijn, L. A. Robertson, M. S. M. Jetten, and J. G. Kuenen. 1996. Autotrophic growth of anaerobic ammonium oxidizing microorganisms in a fluidized bed reactor. Microbiology. 142:2187 ­ 2196.