Development of a Contact Filter for the Removal of Fine Solids From Recirculating Aquaculture Facilities

AIMAP-2009-M16
Kelly Cove Salmon

Table of contents

• Abstract
• Introduction
• Description of the Filter
• Methods
• Results and Discussion
• Conclusion
• Preliminary Recommendations
• References

Abstract

Initial tests performed on the contact filter indicate that the filter is effective at the removal of fine particulates. Overall solids capture efficiencies as high as 51 % have been measured across the filter. The pressure drop across the filter is approximately 440 Pa after cleaning, and rises to about 540 Pa after 36 hours of operation without cleaning. Recommendations are proposed to improve the clean cycle of the filter.

Introduction

Solids removal is an essential process in any recirculating aquaculture system (RAS). Failure to remove the solids, which are generated from fish wastes, uneaten feed, and bio-floc, incurs negative effects on fish health through both direct and indirect means. The direct effects are caused by the irritation of fish gills, whereas the indirect effects include the harbouring of pathogens and the increase in overall oxygen demand in the system (Timmons and Ebling, 2007). With the importance of solids removal stated, it is no wonder that there are several devices employed in a typical RAS for this purpose. For example, at Cooke Aquaculture’s Oakbay Salmon-smolt hatchery facility, where the testing of the prototype contact filter is underway, dual drain fish tanks, swirl separators, and drum filters are employed as methods for solids removal.

However, even these devices working in unison are not 100 % effective, and as a result fine particulates build up in the system. If allowed to accumulate these fine solids will result in the aforementioned problems to fish health, as well as cause problems by plugging system components (Lawson 1995), or supplying a carbon source for heterotrophic bacteria leading to bio-fouling in the bio-filters (Timmons and Ebling, 2007). Currently, the concentration of fine solids in the system is controlled by two processes: the overflow rate of the system and ozonation. The overflow rate of the system is regulated by the inlet of fresh make-up water into the system. This process results in the dilution of the fine particulates, and causes some to exit the system along with the overflow. Unfortunately, the make-up water must be pre-treated before entering the system, and also dilutes the concentration of other desired solutes, which must then be added to the system in order to achieve the appropriate levels. These misfortunes carry with them the burden of increased operational cost. The second method of fine particulate control is ozonation. This method uses ozone to chemically modify the solids thereby making them easier to remove by a separate process, or inactivating them biologically. However, ozone treatment is problematic, since ozone is costly to acquire, and unused residuals of the compound that remain in the water are toxic to the fish (Timmons and Ebling, 2007). As a result of the problems associated with these methods of fine solid control, a cost efficient novel device for the removal of fines is required.

The purpose of this study is to determine if a new type of filter, termed the contact filter, will provide the desired effect of fine particulate removal with relatively low impact on current system processes. The device in question is essentially a depth filter employing an extremely porous bio-carrier media as its packed bed. The advantage of such a device, if effective in the removal of fine solids, is that it should have a relatively low pressure drop removing the need for additional pumping costs. The filter is also designed to be self cleaning removing the need for additional work loads, or labour costs. Cooke Aquaculture has purchased three prototype units which have been installed in parallel in the AB section (one of five sections) of their Oakbay facility.

Description of the Filter

 The prototype contact filter consists of three units, which are each divided by metal screens into three equal sections. A single unit has an overall length of 213 cm, a width of 102 cm, and has a water depth of 123 cm. Each section in a unit is 71 cm long, and contains a packed bed of Curler advanced x-1-135 bio-carrier media. The media is suspended 45 cm above the floor of the filter by a screen so that the media will not drain during cleaning.  Each packed bed of bio-carrier media has an average depth of 56 cm.

Process water enters through the inlet located at the bottom of the unit, flows through the packed bed of media, and exits through the outlet located at the top of the device. The cleaning cycles for the three units are designed to be completely automated and staggered so that one unit is cleaned every eight hours. Each unit is cleaned once every 24 hours. The cleaning cycle begins when the inlet and outlet gates close isolating the unit from the rest of the system. Next, 12-20 cm of water is drained from the filter to prevent transfer of media from one section to another during subsequent steps. Air is then used to agitate the solids off the media, and oscillating water jets are turned on above the media to aid in cleaning. Finally, the unit is drained while the spray jets continue to rinse the media. Once the draining is complete the computer terminal terminates the cleaning cycle by closing the drain gate and opening the inlet and outlet gates.

Methods

After the installation of the filter on May 7, 2010, the commencement of sampling was delayed until May 12 due to a faulty drain valve on the second unit.  Since May 12, sampling has continued on a weekly basis despite minor mechanical problems with the filter, which are to be expected with the development of any new technology. These minor malfunctions are discussed further in the results section.

The sampling for any given week proceeded as follows. Two, approximately 6 L water samples were isolated from the recirculation loop at each of the following locations: the inlet of the contact filter, the outlet of the contact filter, the head tank, and during the cleaning process of the filter. In addition, two 10 L samples of fresh make-up water were also collected. The conductivity of the samples was measured using a Control Company 4075 Wide Range Conductivity Meter. The samples were sealed in their vessels and transported back to the University of New Brunswick, where they were processed for total suspended solids (TSS) and total dissolved solids (TDS) the following day. The flow rate of the make-up water was also determined each week by collecting a known volume of water over a known time interval. The overall capture efficiency of the filter was calculated each week using the TSS data for the samples collected at the inlet and outlet of the filter.

In addition, the pressure drop across the contact filter was measured before and after the cleaning of the filter units by the difference in water heights on the by-pass gate. The system flow rate was also obtained by multiplying the cross sectional area of a region of flow by the water velocity measured using a Nixon Streamflo 430 velocity meter equipped with a 403 Low Speed Probe. The system flow was measured at two locations: at the water weir immediately following the drum filter, and at the exit of each of the three units.

Results and Discussion

The aforementioned mechanical problems associated with the filter included: malfunctions of the drain valves in units 2 and 3, incomplete closing of the inlet gate in the third unit during a cleaning cycle leading to a large loss in system volume, and the overflow of bio-carrier media between sections in the same unit, as well as out of the unit. Due to the problems associated with the gates, the cleaning cycle of the filter is currently operated in manual mode. In addition, the clean procedure for the filter has not been effective at draining all of the solids released from the media during air agitation. Consequently, high concentrations of TSS can be visually observed to rise to the surface of the bed upon restart of the filtration of system water. Further testing will be required to determine the best course of action to improve the cleaning cycle.

The increase in TSS concentrations between week 1 and week 2 can be attributed to a decrease in the make up water flow rate from approximately 400 m3/day in week 1 to approximately 285 m3/day in week 2. The increase in TSS is further explained by an increase in the feed rate between the two dates of 25.5 kg/day.

The filter has proven effective at removing fine solids. The capture efficiency started out at 51 % and has declined slightly over the three weeks of sampling. Further tests will be performed to determine whether the drop in efficiency is due to a change in the size distribution of the solids entering the filter.

The pressure drop across the filter was determined from the difference in the water levels on either side of the by-pass gate. The pressure drop was 539 Pa after approximately 36 hours of operation and decreased to 441 Pa after cleaning. This small pressure drop across the filter was expected since the media employed in the filter forms an extremely porous packed bed, and the bed is less then a meter in depth.

The TDS concentration was measured directly from the weight of dry solids after evaporation of the filtered sample, and indirectly from conductivity measurements. The conductivity method uses a correlation to infer the TDS concentration. The variability in the data can be explained as TDS is a function of the make up water flow rate, the amount of salt added to the system, the bio-mass in the system, and NaOH added to the system to control pH. All of these factors may change on a week to week basis, so variability can be expected. However, the data indicates that the contact filter has little effect on the concentration of TDS.

On May 26, the system flow rate was 0.033 m3/s at the weir immediately after the drum filter and 0.04 m3/s at the outlet of the contact filter. The flow rate should be the same at these two locations when the system is at steady state. The difference between the two values may be due to experimental error or may indicate the system was not at steady state on May 26. Fluctuations in the water level at the inlet of the pump during the cleaning of the filter may affect the flow rate.

Conclusion

The overall solids capture efficiency of the contact filter ranged between 35 and 50 % during the month of May 2010. Additional work is required to improve the cleaning cycle of the filter and to determine the capture efficiency of the filter for different particle sizes.

Preliminary Recommendations 

The technical problems encountered during start up have been minor despite the fact that the filter is a new technology. Several recommendations can be made based on the experience gained so far:

1. Since the filter functions as a depth filtration unit the distance between the screen used to support the media and the floor of the filter should be decreased to allow a greater depth of media.
2. Tests should be performed to determine the effectiveness of the spray jets used in the cleaning procedure. Also, the spray should not be turned on while the cell is being agitated with air because the jets cannot penetrate the surface of the water.
3. Additional testing will be required to determine the best method of refilling a unit after the cleaning procedure to prevent the observed initial break-through of solids upon restart of the unit.
4. A safety mechanism should be installed to trigger a shut down of the units and to open the by-pass gate in the event that a gate or valve become stuck open. This would prevent the loss of system volume should such a malfunction occur.
5. The controller responsible for the cleaning of the filters should be programmed to recognize the time required for cleaning, as to allow the units to be cleaned at the same time each day.
6. The system controlling the inlet and outlet gates of the units should be modified so that the gates can be independently opened and closed in manual mode. 

References

Lawson, T.B., 1995. Fundamentals of Aquacultural Engineering. Chapman and Hall, New York.

Rusten, B., Eikebrokk, B., Ulgenes, Y., Lygren, E., 2006. Design and operations of the Kaldnes moving bed biofilm reactors. Aquacultural Engineering. 34, 322-331.

Timmons, M., Ebeling, J., 2007. Recirculating Aquaculture. Cayuga Aqua Ventures, Ithaca, New York.