CICEET Progress Report for the period 3/15/05 Through 9/15/05

Project Title: Wastewater Treatment to Minimize Nutrient Delivery from Dairy Farms to Receiving Waters
Principal Investigator(s): Katharine F. Knowlton
Additional Investigator(s): Nancy G. Love and Greg M. Mullins
Project Start Date: September 1, 2003 (due to delays in receiving funding, actual start date was December 1, 2003)

Figures

Figure 1a

Figure 1b

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Tables

Table 1

Table 2

Table 3

Table 4

Tasks to meet objectives
Project tasks have been rescheduled, as outlined in the revised project schedule in section III.E. below. We are currently operating the pilot plant system at the Virginia Tech Dairy Center.

After some problems with the simulation package, we recently got the program to compile correctly and have started the simulations. Preliminary results from that effort are included in this report. The EBPR pilot-plant system will be reconfigured this month to operate according to the preferred condition identified by the simulator. We will pilot test one configuration rather than two, due to limited space and fermented waste resources.

As noted in previous reports, start up of the EBPR reactor was the limiting factor during the laboratory study, because the lab reactor was not started with biomass from an EBPR plant. We attempted to avoid the same problem with the pilot-scale system by starting with seed from an EBPR plant in North Carolina; however, we encountered problems with the fermentation process that took time to troubleshoot and are discussed below. Those problems were recently rectified and, fortunately, effective EBPR performance has just kicked in as of the writing o this report. The no cost extension will allow us to complete the project demonstration period.

The rainfall simulation plan for objective 2 will be conducted in the next month if the simulated EBPR configuration can be successfully operated in a way that removes P. We hope those studies and all analyses can be completed by November, and a final report will be generated by the end of December.

Progress on Tasks
Task 1.1 Fermentation and wastewater analysis.
The laboratory phase of this experiment is complete and discussed in section II.

Task 1.2 Parameter estimation.
The parameter estimation exercise is complete and presented in section II.

Task 1.3 Simulate optimum operating performance of EBPR system.
The model simulations that evaluate optimum operating configuration for the EBPR system are underway, and preliminary results are presented in section II. We have shifted from using BioWin to AQUASIM, a simulation package that provides much greater flexibility in the modeling exercise than BioWin.

Task 1.4 Construct pilot-scale fermentor and EBPR system at Virginia Tech Dairy.
This task is complete.

Task 1.5 Operate and evaluate the pilot-scale treatment system.
The fermentor is operating, after some major modifications that are discussed in section II. The EBPR reactor is operational at the moment but not yet optimal. We expect the reactor to be optimal by October.

Task 2.0 Perform rainfall runoff simulation studies of both the wastewater and biosolids generated by the pilot plant system.
This task will be conducted in November 2005 if we can get the EBPR system sufficiently stable before cold weather sets in.

Difficulties
We encountered difficulties in the operation of the laboratory-scale EBPR reactor. These difficulties related to the initiation of nitrification, which reduces the efficiency of EBPR. Many studies of EBPR on agricultural wastes in the literature add nitrification inhibitors to eliminate this potential. We did not want to add an inhibitor, because it would never be done in a full-scale facility. We implemented step feed early in the project to prevent nitrification in the bioreactor, but continued to have problems, so we adjusted the cycle times with the step feed. This eventually minimized the degree to which nitrification occurred and efficient EBPR was achieved. Once achieved, the parameter estimation experiments were reinitiated.

Another difficulty we encountered was the lack of volatile fatty acid (VFA) production potential in the pilot-plant influent. This resulted in poor P removal and we ended up investigating the reasons for this with two fermentation potential studies, which are presented in Section II. From the fermentation potential studies, we learned that the wastewater generated from fresh dairy manure used during the lab-scale system had substantially higher fermentation potential than the wastewater generated from the fresh dairy barn manure that is flushed into the full-scale treatment system and feeds the pilot-plant. We have sufficient buckets from the pilot plant study in storage to complete this study, so the manure source was changed to the material collected during the laboratory-scale study. Once the wastewater was changed, the amount of VFAs in the EBPR reactor feed increased and, shortly thereafter, the EBPR process started to work.

Project Objectives for Next Reporting Period

Objectives
By the next reporting period, we anticipate having achieved all project objectives.

Tasks to Meet Objectives
The primary tasks to be completed are tasks 1.5 and 2.0, as described in Section I.C.

Work Plan for Next Reporting Period
The pilot-plant system is functioning. Our final tasks are to continue to monitor it while operated under the optimal configuration, as determined by the simulation. Then, the same six stoichiometric and kinetic parameters that were estimated during the lab study will be estimated from the biomass in the pilot plant. Additionally, effluent from the EBPR system will be applied to York River Basin soil to perform a rain simulation and to estimate runoff potential of the effluent phosphorus after EBPR treatment.

Anticipated Success in Meeting Project Objectives
We anticipate being able to meet all project objectives.

Overall Project Timeline Update
The overall project timeline is given in Table 4, and shows that the project should be completed by December 2005; however, if the timing of achieving EBPR is delayed so that the rainfall simulation has to be delayed, pilot plant effluent will be frozen and stored for evaluation in the spring.

Preliminary Data
A. Dairy Wastewater Composition
The composition of separated dairy manure was reported in an earlier report. Six early lactation cows were fed a common diet, and all excreted urine, feces, and milk were collected on 3 consecutive days for three periods (switchback design). All excreted urine and feces were mixed by cow in the proportions excreted. Water was added to create a diluted slurry (50% water w/w) to reflect the common practice of flush removal of waste from the barn. Water, feces, and urine were mixed for three minutes with a paint mixing paddle attached to a drill. Slurry from each cow was stored at ambient temperature for 24 h, and then mixed, subsampled and pumped through a mechanical solids separator. The separator consisted of two perforated basins (concave screens), with a rotating brush assembly to convey slurry across the screens. Liquids flow through the screen to a collection basin below the screens. Solids are advanced and forced out of the discharge by the rotating brushes. Liquid effluent and manure solids were weighed and subsampled, then analyzed in triplicate. The liquid effluent was used to feed the laboratory-scale fermentor and EBPR reactor. It is currently being used to feed the pilot-plant due to problems encounted with the full-scale manure, which is contaminated with high levels of inert suspended solids that significantly reduce the fermentation potential (discussed below).

B. Laboratory-Scale Reactor Performance and Parameter Estimation
The overall performance of the laboratory-scale fermentor and EBPR reactor were shown in a previous report. The fermentor produced a high VFA content effluent. These were essential for successful EBPR. The EBPR reactor achieved EBPR and the biomass from this reactor was used for parameter estimation studies. The parameters estimated were given in the previous report. The parameter values available in the literature cover a broad range. The parameter estimates we derived are generally in the range obtained in the literature for other wastewaters.

With the EBPR underway, the soluble reactive phosphorus usage of the reactor was higher than what would have been expected for biomass growth alone. Except for the times when nitrification occurred in the system, soluble reactive phosphorus concentrations of the effluent were around 0.3 mg P/L. The system ran at P-limited conditions and reached %P values of 2-3% in biomass. The average percent effluent soluble reactive phosphorus removal achieved by the reactor during the kinetic experiments was 85%. The suspended solids concentrations in the effluent towards the end of the reactor operation were higher than expected, yet the reactor still ran at P-limited conditions.

C. Pilot-Scale EBPR System and Fermentor at the Virginia Tech Dairy
The pilot scale fermentor system and pilot-scale EBPR reactor have been monitored since March 7, 2005 and March 31, 2005, respectively. Figure 1A and Figure 1B show a schematic representation of the pilot-scale systems used in this phase, and Figure 2 is a picture of the actual system.

C.1 Reactor Suspended Solids Stability
The data show that the total suspended solids (TSS) concentration in the pilot-scale fermentor and EBPR system changed over time (Figure 3). A biomass (mixed-liquor) TSS concentration of less than 5,000 mg/ L TSS is necessary to achieve biological treatment that includes an aerobic phase, as is true in EBPR. In wastes with higher TSS concentrations, it is difficult to supply enough dissolved oxygen to the reactor. This cutoff is both an accepted norm in the design process and a lesson-learned of the lab-scale system we ran at our labs for the parameter estimation portion of this study.

The field-scale settling basin initially served as our source of solids-separated wastewater for treatment by EBPR, to mimic that which was done manually during the laboratory-scale study. The settling basin receives post-solids separated dairy manure and in theory, wastewater here should show similar properties to the wastewater used in the lab-scale system, which also was solids-separated through a pilot-scale solids separator at the Virginia Tech Dairy. Furthermore, with preliminary experiments, we determined that the front part of the settling basin has the highest amount of VFAs (the organic acids essential to achieve successful EBPR; data not shown). For this reason, we chose the front end of the settling basin as the source of wastewater for the fermentor.

As shown in Figure 3 the TSS concentration of the settling basin varied between 10,000 and 27,000 mg/L TSS. The settling basin feed was fed into the fermentor and then the fermentor effluent was fed into the EBPR reactor to achieve EBPR. These high TSS values in the basin resulted in rather high biomass (mixed liquor suspended solids, or MLSS) concentrations, stabilizing around 20,000 mg/L TSS in the system during the first two weeks of the study. In order to bring this value down to more acceptable limits, we employed a 1:2 dilution of the settling basin water fed into the fermentor. This brought the MLSS concentration down to about 10,000 mg/L TSS, which later stabilized around 15,000 mg/L TSS between days 48 to 57. In order to bring these values down more, we employed more aggressive dilution, resulting in a final dilution of 1:10 to 1:15, depending on pump performance. As seen in Figure 3, this new dilution rate resulted in a more desired TSS concentration within the biological reactor over time and as of operation day 147, it stabilized around 3,800 to 5,000 mg/L TSS. The effluent TSS also stabilized at low concentrations of about 110 mg/L TSS.

C.2 Fermentation Potential Analysis
The field-scale treatment system did not perform EBPR at the start of operation. A primary cause was thought to be the low VFA concentration in the feed and/or improper fermentation. Initial fermentation potential analysis confirmed that the fermentor did not ferment. We know that the system had high concentrations of volatile suspended solids (VSS) which should have allowed fermentation, but more detailed fermentation potential analysis were conducted to compare the fermentation potential of three potential feed streams. The first feed stream analyzed was the wastewater used in the parameter estimation studies performed with the lab-scale system (referred to hereafter as lab-scale feed). This feed stream is the wastewater collected during the cow study during the first phase of the project. The second feed stream analyzed was the wastewater from the settling basin (SB). This is the feed that we used at the startup of the unsuccessful pilot-scale system at the Virginia Tech Dairy Center. The third feed stream analyzed was primary sludge (PS) obtained from the Blacksburg Wastewater Treatment Plant (WWTP). This feed should inherently have a very high fermentation potential. In addition, a fourth test was conducted, which consisted of a blend of PS (10%) and SB (90%). The fermentation potential of these four feed stream combinations were analyzed, and the results are summarized in Table 1 and Figure 4.

The data show a significant increase over time in the total VFA (TVFA) concentrations in PS and lab-scale feed primary sludges, but substantially less in the SB waste. The VFAs in the PS, which served in effect as a positive control for the fermentation potential assay, increased from 0.09 to 0.21 mg VFA as COD/mg initial VSS over a course of 8 days. The lab-scale feed had the highest fermentation potential result and increased from around 0.16 to 0.73 mg VFA as COD/mg VSS over 8 days. One hundred percent (100%) SB wastewater (used as feed into the pilot-scale fermentor) had the lowest TVFA concentration of all four potential assays, with a change from 0.009 to 0.015 mg VFA as COD/mg VSS, and demonstrated peak VFA formation potential at day 6 with 400 mg COD/L, of which about half was acetic acid. Acetic acid is the most essential of VFAs for successful EBPR. The 10% PS and 90% SB wastewater mixture resulted in better fermentation, but still not enough VFAs to satisfy the need for the downstream EBPR system. This problem of insufficient VFA was made worse by the significant, aforementioned dilutions we have to perform to keep the MLSS concentration in the bioreactor low enough to supply enough dissolved oxygen in the aerobic zones. The data in Table 1 largely corroborate the VFA data in that a significant loss of VFA over the course of a fermentation potential test typically translates into high VFA formation.

To validate the fermentation potential results, we repeated the study. Wastewater samples were taken at pre and post solids separated locations from the full-scale dairy facility for comparison with a fermentation potential test. This time, the time length of the test was extended to 12 days. The results show that the fermentation potential of pre and post solids separated wastewaters at the dairy facility were similar (Figure 5). The pre-solids separated wastewater was slightly more fermentable than the wastewater following separation, but was still much less fermentable than the lab-scale feed. In addition, the data indicate that the VSS did not decrease over the course of the 12 day experiment (Table 3). This is consistent with the low VFA potential. Finally, we analyzed methane (CH₄) and carbon dioxide (CO₂) production during the fermentation potential test for both wastes throughout the 12 days of the study. Production of these gasses was similar (Figure 6) and suggests that the cultures were primarily performing methanogenesis (converting organic waste materials into methane) rather than fermentation (converting organic waste materials into VFA).

Overall, the difference between the fermentation potentials of the SB versus the lab-scale feed was clear, and prompted us to begin using the lab-scale feed for the pilot scale system. This introduced some operational difficulty because the lab-scale feed is only available in small quantities and must be hauled to the site daily. To accommodate this, we had to down-size the fermentor; however, this change resulted in proper fermentation. TVFA concentrations of 5,000 mg/L COD in the fermentor effluent have been obtained recently and the fermentation efficiency is expected to increase to the values seen during the lab-scale study (Figure 7). This is a critical element needed to fuel EBPR.

We performed an analysis to determine why the high VSS wastewater from the settling basin had such a low fermentation potential (Table 1). We suspected that the poor fermentation potential of the settling basin wastewater was primarily due to the high fraction of inert organic suspended solids in the system (inert VSS). Table 1 shows that the VSS concentrations of the primary sludge and lab-scale feed declined over the 8 day fermentation potential test period. These declines indicate excellent fermentation. On the other hand, the settling basin feed and the blend of 10% PS and 90% SB feed did not show this characteristic decline. This confirmed that the inert organic suspended solids fraction, which does not contribute to fermentation, of the settling basin feed was quite high causing poor fermentation potential.

This series of fermentation potential studies made clear the need to determine if the very high fraction of inert organic matter in the settling basin wastewater was an unexpected consequence of the normal operating strategy in place at the full-scale dairy center. An alternative explanation is that it was due to a malfunction in the full-scale solids-separator, causing it to remove mostly biodegradable (non-inert) organic matter into the solid waste it generated, making the downstream liquid manure stream low in biodegradable organic matter. To evaluate the cause, a TSS balance on the whole Virginia Tech Dairy Center was calculated. It was found that almost 60% of the water that goes through the separator was actually “grey water” used for flushing the barn for cleaning purposes 4 times a day (Table 2). This “grey water” is circulated through the system in order to conserve water. This is a common practice in full-scale systems since it generates lower wastewater volumes and saves on fresh water use by the industry. While this grey water is relatively low in TSS, its high flow rate makes it a significant contributor of TSS. The problem with this grey water is that due to its recirculation, the organic matter found in it has lost most of its biodegradable components through biochemical processes. What is left is essentially inert, and does not support fermentation and biological treatment.

These results indicate that the most critical operations for biological dairy manure treatment are the ones that are designed to reduce the amount of total solids. Properly designed systems can be efficient enough to reduce the solids that are introduced into the treatment stream via grey water.

C.3 Pilot-Scale System Performance Data
Recent changes to the pilot-scale system have improved overall performance. The data show that changing to lab-scale feed and diluting it appropriately resulted in TSS concentrations in the EBPR reactor of between 4,000 and 5,000 mg TSS/L (Figure 3). The effluent TSS has gradually decreased to about 100 mg/L TSS, which is reflective of good settling for this type of treatment system. The total phosphorus (TP) removal has been improving over the last two weeks once efficient VFA formation was achieved, and was above 90% (effluent soluble P < 0.3 mg/L) for the last point measured (Figure 8). Figure 9 shows the change in soluble reactive phosphorus (SRP) concentration over time. As can be seen, we have recently begun to realize effective EBPR (after day 150). This is excellent and allows us to complete the final tasks of the project.

One of the key parameters that indicates the occurrence of the EBPR metabolic processes is the %P in biomass. Figure 10 shows the change in this parameter with respect to time in the field scale system. Biomass P content greater than 2% is an indicator of the probable presence of phosphorus accumulating organisms (PAOs). Currently, this value is at 2.1%.We anticipate further improvements as the system continues to stabilize.

D. Simulation Studies
The simulation studies are currently underway and will be completed within a week of this report in order to define the optimal reactor configuration for EBPR. Preliminary data are available and shown here. Activated Sludge Model 2d (ASM2d) has been simulated using a software package called AQUASIM, produced and distributed by EAWAG. The complex nature and over parameterization of ASM2d made its implementation in AQUASIM more difficult than anticipated. However, those challenges were overcome and the simulation package is working well now.

Different solids retention times (SRTs) were simulated to determine the optimum SRT that results in the lowest soluble phosphorus concentration in the effluent. Through simulation and as show in Figure 11, this SRT was found to be 4.5 days. Our base value of 4.27 days of total SRT in the lab-scale system is close to the optimal value found with the simulation studies.

As mentioned in earlier reports, EBPR is accomplished by moving phosphorus accumulating organisms (PAOs) between anaerobic and aerobic zones/reactors. For this reason, optimal exposure time of PAOs to these different environments has also been investigated in the simulations while, at the same time, preventing nitrification. We found in the simulation studies that the optimal anaerobic:aerobic reaction ratio is 0.375 (Figure 12). The ratio we used in the lab-scale system was longer at 0.57, but may have been necessary based on factors that were not simulated (oxygenation efficiency in a small scale system, for example).

Dissemination
Workshops:
Results of this project were presented at the Virginia Tech Feed and Nutritional Management Cow College in Blacksburg, VA in January, 2005. Approximately 75 dairy industry consultants were present, and the project was well received.

Conferences:
5 total (4 presented, 1 pending)
Presented:
K. F. Knowlton, N. G. Love and M. Muftugil. Wastewater treatment to reduce the phosphorus content of dairy manure. pp. 129-134 In Proceedings of the Virginia Tech Cow College, Blacksburg, VA. January 2005.

Weaver, C. C., Muftugil, M. B., Kozarek, J., Wolfe, M. L., Knowlton, K. F. and Love, N. G. Utilization of a Fermentor to Support Enhanced Biological Phosphorus Removal with Dairy Wastewater. Presentation at Undergraduate Research Day, Virginia Tech, October 14, 2004.

Muftugil, M. B., Love, N. G. and Knowlton, K. F. Using Enhanced Biological Phosphorus Removal (EBPR) to Alter the Nitrogen:Phosphorus Ratio of Dairy Manure and to Minimize Nutrient Delivery to Receiving Waters. Presentation at Innovative Uses of Agricultural Animal Manure, Biosolids and Paper Mill Residuals, June 29 ­ July 1, 2005 Omni Chicago Hotel, Chicago, Illinois USA.

Muftugil, M. B., Love, N. G. and Knowlton, K. F. Using Enhanced Biological Phosphorus Removal to Minimize Nutrient Delivery from Dairy Farms to Receiving Streams. Submitted to the Virginia Water Environment Association/American Waterworks Association Joint Annual Meeting, Virginia Beach, Virginia, September 2005.

Scheduled Presentations:
Muftugil, M. B., Love, N. G. and Knowlton, K. F. Using Enhanced Biological Phosphorus Removal to Minimize Nutrient Delivery from Dairy Farms to Receiving Streams. Submitted to the Water Environment Federation WEFTEC 2005, Washington DC, October 2005.

Expenditures
Approximately 92% of funds are expended. We have sufficient funds to complete the project by the end of the calendar year, assuming the soil leaching test can be conducted before the end of the growing season.