The Long Path, Plug Flow, Submerged Biological Filter

Introduction

Operators of recirculating aquaculture systems are blessed with a wide variety of choices when it comes to biological filters.  Most of the filters used are either well mixed reactors with medium to long retention times or short path plug flow reactors with relatively short retention times.  These systems are sufficiently efficient and effective when used for raising crops like tilapia that are tolerant of poor water quality.  The problems start when farmers try to raise crops that are not tolerant of poor water quality. The biofilters commonly used today are very inefficient when very low levels of ammonia are required.    

The tough part about maintaining the water quality necessary to raise species sensitive to ammonia is doing it while feeding the fish at the levels necessary to produce a crop in an economic time frame.  It is not that difficult to maintain water quality while feeding at 0.5% of body weight in a system stocked at 0.1lb of fish/gal of water.  It is entirely different to maintain water quality while feeding at 3% of body weight in a system stocked at 0.75lb/gal.  It is the difference between a hobby aquarium and a profitable recirculating aquaculture system.   

For the foreseeable future, the most intelligent solution for biofiltration and suspended solids control in recirculating systems will be to employ several different types of water quality control devices simultaneously.  Each device should have one primary function. This has numerous advantages.

1. Each individual device can be simpler.
2. Simpler devices tend to be more reliable.
3. In the event of a malfunction or failure of a single device, the overall effect on the system will not be as great. In some cases, other components of the system can compensate for the loss of one.
4. Downtime for maintenance of one component in a multi-component system has less of an effect than losing the “do it all” type biofilter/solids.
5. If a multi-component strategy is adopted, it is easier to add more components as the need arises.
6. Each type of biofilter or solids control device has its own characteristic advantages and weaknesses. By using several different devices simultaneously, these devices can compliment each other and do a better overall job of maintaining water quality.
7. Using numerous devices increases the flexibility of the overall system and enables the farmer to change crops as the market dictates the need.

Types of reactors and their characteristics

There are two basic types of biological and chemical reactors: Continuously Stirred Tank Reactors (CSTR) and Plug Flow Reactors (PFR).  CSTR are also known as Well-Mixed Reactors. CSTR or Well-Mixed Reactors are easily visualized as vessels or tanks that are stirred to achieve uniformity throughout the tank. 

Figure 1: Well-Mixed (Continuously Stirred Tank) Reactor

A very important characteristic of the CSTR is that the concentration of the reactants in the outlet is equal to the concentration of the reactants in the vessel regardless of the concentration of the reactants in the inlet.

Figure 2: Plug Flow Reactor

A PFR can be modeled as a pipe where the reactants move as a plug through the pipe.  The concentrations of the reactants in plug flow reactor will vary along the pipe and there is no mixing between the beginning and the end of the system.

One of the obstacles facing recirculating systems trying to culture species with high water quality requirements is the current reliance on CSTR bioreactors or Short Path Plug Flow Reactor (SPPFR). Here are some examples of these types of systems. 

• Moving Bed Biofilter (MBB) – This type of biofilter is a good example of a CSTR.  It adds oxygen to the water during normal operation and it has a very low operating head.  A number of small MBB’s arranged in series could approximate a long path plug flow reactor.
• Fluidized Bed Sand Filters (FBSF) – These filters are half-way between CSTR and PFR.  If a system were designed with numerous smaller reactors in series, the system may approach the operating characteristics of a PFR.  However, the high pressure drop of FBSF’s make this possibility economically impractical. 
• Bead Filters – These are primarily SPPFR.  Unfortunately, they suffer from high pressure drop and cyclical performance as they load up with solids and require backwashing.  The major problem with bead filters is the conflicting and mutually exclusive tasks of collecting solids and removing ammonia. The collection of solids encourages the growth of heterotrophic bacteria at the expense of nitrifying bacteria.
• Trickling Filters – These are very good PRR’s, but the path is short.  The short path drawback can be overcome by using several trickling filters in series, but the energy required for all the pumping would make it uneconomical.  Trickling filters are the best all-around biofilters due to their ability to do gas exchange and biofiltration.  Their only real drawback is the pump head required to lift the water up to the top of the filter.
• Rotating Biological Contactors (RBC) – RBC’s have the potential to approximate a long path, plug flow reactor if multiple units are arranged in series.  Since RBC’s have very low pump head requirements it is very feasible to use them as an LPPFR. 
• Submerged Bed Packed Filters – These types of filters operate best as Long Path, Plug Flow Reactors (LPPFR).  They have very low pump head requirements and operate hydraulically better with a long flow path and long retention times.  Traditionally, submerged filters have been up-flow or down-flow.  However, horizontal-flow is the best orientation for an LPPFR.   

Comparison of CSTR vs. PFR

If the species being cultured can tolerate fairly high concentrations of ammonia, a CSTR can be an excellent choice for a biofilter.  Above 2.5ppm TAN (Total Ammonia Nitrogen), ammonia removal rates are constant and independent of ammonia concentration.  Therefore, there is no difference in overall efficiency between well–mixed reactors and PFR if the desired outlet concentration of the biofilter is 2.5ppm TAN or greater. 

Figure 3: Total Ammonia Nitrogen (TAN) Removal Rate

Below about 2.5ppm TAN, the ammonia removal rate starts to become dependent upon the ammonia concentration. The rate at which biofilters remove ammonia is proportional to the square root of the concentration of ammonia. This is known as a 1/2 order reaction rate.  The following graph shows a typical ammonia removal curve. 

In a Well-Stirred Reactor, the influent water is quickly diluted so that the ammonia concentration is equal to the outlet concentration.  Thus, if the inlet concentration is 1.0ppm TAN and the desired outlet concentration is 0.1ppm TAN, the bioreactor must be sized based on the very low removal efficiency at 0.1ppm.   

In contrast to a CSTR, a PFR starts off with a removal efficiency that corresponds to the inlet concentration.  The efficiency decreases through the reactor as ammonia is removed and the concentration drops.  Only at the very end of the flow path will the efficiency of ammonia removal be down to the removal efficiency of the whole Well-Mixed Reactor. 

Figure 4: CSTR/LPPFR Surface Area Ratio vs Ammonia Removal Efficiency

One way to look at the advantage of an LPPFR versus a CSTR is to look at the amount of surface area required to accomplish a given ammonia removal rate.  The following graph shows the increasing advantage of the LPPFR as the fraction of ammonia removed increases. 

If the cultured species is sensitive to ammonia and low concentrations are desired, an LPPFR can be significantly more efficient than a CSTR. 

Another advantage of an LPPFR, is the establishment of sequential zones with different bacteria predominating.  The inlet section of the biofilter will tend to have heterotrophic bacteria.  After the Carbonaceous Oxygen Demand (COD) has been absorbed, nitrosomonas bacteria will be able to establish themselves and oxidize the ammonia to nitrite.  The production of nitrite will allow the nitrospira to establish themselves as a third zone.  These zones will not be sharply delineated but will gradually progress from one to another. 

Physical Description

In the past, a common way to construct submerged filters was to build large flat beds of media with water flow either up–flow or down–flow.  This is the worst way to build a submerged filter.   It virtually guarantees plugging, channeling, short circuiting of water, and overall poor performance. 

Figure 5: Up-flow or Down-flow Submerged Filter Configuration

What does a Long Path, Plug Flow Bioreactor look like?  At the present time, the most practical method of constructing a long path, plug flow system is to use a horizontal–flow, submerged filter.  The construction of tall, narrow, vertical columns is an expensive way to build biofilters.  Horizontal raceways are much easier to build and are less expensive.  There are several possible configurations for a submerged filter operated as an LPPFR. 

• The simplest is a long raceway.

Figure 6: Long Horizontal Raceway LPPFR

• The next obvious possibility is a folded raceway.

Figure 7: Folded Raceway LPPFR

• A tank can be baffled any number of times to achieve the desired flow path.

Figure 8: Multiple Baffle/Folded Raceway

• Although it is not optimum in terms of tank utilization, a folded vertical system is also an option. 

Figure 9: Vertical-Flow Multiple Baffle/Folded Raceway

All of the above examples assume a rectangular tank.  Circular tanks can also be used.  By using concentric tanks, an annulus can be created that is an excellent LPPFR.  Starting with a 12ft (3.7m) diameter tank, the length of the filter would be almost 38ft (11.6m) long.  For a 20ft (6.1m) diameter tank the length would be almost 63ft (19.2m).

Figure 10: Circular LPPFR – with the biofilter in the outer ring

Pressure drop through an LPPFR using structured media is very low.  The current trend toward low head systems is easily satisfied by an LPPFR.  The graph below shows head loss at different superficial face velocities.  These results are for clean media.  Biofilms on the surface of the media will increase the resistance to flow but the total pressure drop will still be very low.  Based on observations and measurements, system designs are targeted to velocities between 4–10ft/min (1.2–3.0m/min). 

Figure 11: Pressure Drop Through Submerged Media

Besides the energy savings, another benefit to a low head system is the ability to accommodate a large flow of water.  A low turnover time for the culture tank is an important factor when trying to keep ammonia concentrations low. 

Low head systems often use air lift pumps to move water. One of the additional possibilities to increase the flow rate through the system and accommodate higher head losses is to use multiple air lifts at different stages of the reactor.  The additional aeration is another benefit.  Since nitrification rates are also dependent on oxygen concentrations, intermediate air lifts can add to the overall efficiency and ensure high oxygen concentrations when the water is returned to the culture tank. 

The key design feature of an LPPFR that ensures plug flow is a sufficiently high velocity.  This creates turbulent flow and prevents back mixing.  Creating a long thin channel for water flow has a number of practical benefits: 

1. High water velocity ensures plug flow. 
2. Ammonia removal rates are proportional to Reynolds number so higher water velocities mean higher ammonia removal rates. 
3. Plug flow means that retention times are uniform, and all of the water is treated for the same amount of time. 
4. There are no complicated measures necessary to achieve an even water distribution. 
5. The design is both flexible and simple. 

Media Selection

One of the advantages and disadvantages of an LPPFR using structured media is the ability of the biofilter to remove very small particles.  Biofilms are very “sticky” and tend to trap small particles.  For this reason, correct selection of the media is very critical.  Ideally, the media will continue to work regardless of the solids collected.  A high void fraction and large free passage diameter is essential, which means that gravel, fiber pads, and ribbon bundles are not recommended.  Random dumped plastic media can be used, but the high cost and difficulty of handling random dumped media make it a poor choice.  The best type of media is structured media with sheets or plates that are parallel to the water flow direction.  This type of media is critical to avoid plugging and to keep the head loss low.  A typical example of a cross–corrugated structured media is shown here. 

Figure 12: Typical Cross-Corrugated Structured Media

When sizing a biofilter, it is always advantageous to use a media with the most surface area per unit volume, provided that it will not plug.  However, there is no rule that requires the use of only one type of media.  Varying the media density through the biofilter helps to minimize both the size of the biolfilter and the maintenance requirements.  The inlet section should have the most open or least dense media, and following sections can use progressively denser media. 

Summary

A Long Path Plug Flow Bioreactor (LPPFR) using structured media has a number of significant advantages.

1. Low pump head loss minimizes energy costs.   
2. Low pumping costs allow high volume flows and fast tank turnover times.  
3. An LPPFR reactor requires less surface area than a CSTR when trying to achieve low ammonia concentration levels. 
4. Plug flow ensures consistent treatment of all of the water passing though the biofilter. 
5. Easily removed packing allows convenient observation of operating conditions within the biofilter. 
6. Flexible designs can accommodate various farm layouts. 
7. An LPPFR can collect very fine solids that are normally difficult to remove. 
8. An LPPFR has the ability to efficiently remove ammonia to very low levels. 
9. An LPPFR can accommodate variable water flows and nutrient loadings.  
10. Variable operating conditions allow the flexibility to change crops to accommodate market demands. 

References

1. Greiner, A. D., Timmons, M. B., 1998. Evaluation of the Nitrification Rates of Microbead and Trickling Filters in an Intensive Recirculating Tilapia Production Facility. Aquacultural Engineering pp.189-200. 
2. Kamstra, A., Van der Heul, J.W., Nijhof, M., 1998. Performance and Optimization of Trickling Filters on Eel Farms. Aquacultural Engineering pp.175-192. 
3. Saucier, B., Chen, S., Zhu, S., “Nitrification Potential and Oxygen Limitation in Biofilters” presented at the Third International Conference on Recirculating Aquaculture July 2000. 
4. Timmons, M.B., Losordo, T. M., 1994. Aquaculture Water Reuse Systems: Engineering Design and Management. Elsevier Science B.V. 
5. Zhu and Chen, 1999. An Experimental Study on Nitrification Biofilms Performances Using a Series Reactor System. Aquacultural Engineering, Vol 23, pp.245-259.   

Published by L S Enterprise LLC
P.O. Box 261
East Earl, PA 17519-0261 USA
Author: Matt Smith
Tel#     239-851-1175
Email: matt.smith@lsentllc.com

Rev. 1/2/2025

©1995-2025 by L S Enterprise LLC. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.