Sizing Biofilters - Part II.

By Sergio Chaim

When I began this series on biofilter sizing I already knew that many breeders among us are skeptical about this technology, so intended to do not allow my efforts become a complete wasting of time I planed write this series in way that it could be helpful even for the skeptical ones who yet use regular hose and bucket system or those who use a more advanced water change system but do not yet perform water recycle. 

"In recirculating systems, good water quality must be maintained for maximum fish growth and for optimum effectiveness of bacteria in the biofilter.   Water quality factors that must be monitored and/or controlled include temperature, dissolved oxygen, carbon dioxide, pH, ammonia, nitrite and solids. Other water quality factors that should be considered are alkalinity, nitrate and chloride." Masser et al. (no date). Sumerfelt and Losordo both in Ebeling (no date) designed biofilters for cold-water and warm-water species, respectively, based only on dissolved oxygen demand and TAN production. Wheaton (no date) did the same for trout,  Hochheimer and Wheaton (no date) too. 

For now I'll discuss about the factors which would guide a rational water  management based on Colt (1986) proposal for Mass Balance Approach. He proposed to design flow-thought systems for salmonids culture  taking into account dissolved oxygen, carbon dioxide, ammonia, and fecal solids production as possible limiting factors. In short, the question he proposed is to identify what water parameters are limiting fish welfare and get rid of it performing water changes. 

I'll discuss about biofilter sizing itself in a future Part III.

4 - Water Change.

4.1 - Days Between Changes.

This is the interval between changes, it is expressed as days (d).

Of course frequent changes are better for the fishes than spaced ones, but most among us have other obligations than care the fishes. Wait too much to perform water changes is not only bad due accumulation of higher amounts of waste or possible depletion of oxygen in systems where there is no aeration, but because sudden changes in water parameters can be disastrous too. 

I'll assume daily water changes.

4.2 - Maximum Nitrogenous Waste Level.

This is the maximum concentration of nitrogenous waste, as nitrogen basis (ammonia-nitrogen, nitrite-nitrogen or nitrate-nitrogen) desired to be present in the water, it is expressed as milligram per liter (mg/l).  

The nitrogen originated from uneaten feed, fish excreta, organic debris and atmosphere are found at aquacultural environment  in organic and inorganic forms. These organic forms are readily converted to inorganic ones, mainly ammonia, and ammonia is converted to nitrite and nitrite to nitrate. Here our main concern is to design a filtering unit to transform potentially high toxic ammonia to relatively "safe" nitrate and to get rid nitrate with water changes, although there are technologies which allow us get rid of nitrate by other ways that not wasting water (denitrification).  

In a regular  hose and bucket system you would decrease ammonia level, but in a recycle system because ammonia is converted to nitrate by bacterias living attached in the filter and since fishes tolerate higher levels of nitrate than ammonia recycle systems demand smaller water changes than non-recycle ones. If properly operated, after filter start-up phase, should there is no significant accumulation of nitrogenous waste other than nitrate in a recycle system. 

Un-ionized ammonia nitrogen (NH3-N or UIA), ionized ammonia nitrogen (NH4-N), nitrite nitrogen (NO2-N) and nitrate nitrogen (NO3-N) are converted to their original amounts (non-nitrogen basis - un-ionized ammonia (NH3), ionized ammonia (NH4), nitrite (NO2) and nitrate (NO3)) multiplying them by 1.21, 1.29, 3.29 and 4.43, respectively. 

Ammonia Toxicity.

Ammonia is present in the water in two forms, un-ionized ammonia and ionized ammonia, with their relative proportions being a function of pH, temperature and salinity. While increases in pH and temperature increases the proportional amount of the most harmful  NH3, raising salinity you decrease relative amount of NH3. An un-ionized ammonia calculator is available from Aquanic. 

 FAO-WHO (1986) cites: “Susceptibility of mosquito fish to ammonia was studied by Hemens (1966) who reported a 17-h LC50 of 1.3 mg NH3/liter; he also observed that male fish were more susceptible than females. Rubin & Elmaraghy (1976, 1977) tested guppy,  Poecilia reticulata, fry and reported 96-h LC50 values averaging 1.50 mg NH3/liter; mature guppy males were more tolerant, with 100% survival for 96 h at concentrations of 0.17 - 1.58 mg NH3/liter.” and “Tabata (1962) conducted 24-h tests on the toxicity of ammonia for  Daphnia (species not specified) and guppy at different pH values and calculated the relative toxicity of NH3/NH4 to be 48 for  Daphnia and 190 for guppy (i.e., NH3 is 190 times more toxic than NH4).” LC50 is the amount of a substance needed to kill 50% of test animals, for ammonia it is usually evaluated for 96 hours exposure. 

 US EPA (1999)   ,take care it is a large .pdf file, is an extensive study intended to establish water quality standards to protect aquatic life from acute (short-term) and chronic (long-term) effects of ammonia exposure. They presented Figure 1 that was adapted from Tabata (1962), and  Tables 1  and  2 which summarize their recommendations in terms of  TAN: 

Figure 1 - Effect of pH and TAN concentrations on short-term ammonia toxicity for guppies. 

From US EPA (1999) .

 

   

 

Table 1 - US EPA recommendation of maximum TAN levels to avoid short-term toxicity for aquatic life.

 From US EPA (1999). 

 

 

Table 2 - US EPA recommendation of maximum TAN levels to avoid long-term toxicity for aquatic life. 

From US EPA (1999). 

 

Searching my papers I found a few more numbers:

Ebeling (no date): "Un-ionized ammonia concentrations below 0.05 mg/l are recommended for long-tern exposure".

Hochheimer and Wheaton (no date) cited  0.0125 mg un-ionized ammonia nitrogen per liter as a requirement of striped bass.

Tucker (1993): "Suggested criteria for environmental ammonia are usually based on the concentration of un-ionized ammonia, and range from less than 0.01 to more than 0.2 mg/l NH3-N (Meade, 1985). A maximum concentration of 0.0125 mg/l NH3-N has been suggested for optimum health in salmonids under continuous exposure (Smith and Piper, 1975). However 0.04 mg/l NH3-N is considered "reasonably safe" for production of rainbow trout in raceway systems (Meade, 1985). Warm-water fish are somewhat more tolerant to ammonia than cold-water fish (Arthur et al., 1987), but insufficient date are available on which to base recommendations for "safe" concentrations. The criterion set by the United States Environmental Protection Agency (1976) as "safe" for freshwater aquatic life under continuous exposure is 0.02 mg/l un-ionized ammonia-nitrogen."

Noga (1995): "Unionized ammonia levels greater than ~1.00 to 2.00 mg/l are usually lethal within 1 to 4 days (Meade, 1985). Below this level fish might not die, but they will be stressed. If UIA is greater than 0.05 mg/l, it should be reduced as quickly as possible." 

Losordo and Hobbs (2000) suggested that tilapias could be raised in water containing 1.8 mg TAN per liter.

Buttner et al. (no date) mentioned less than 0.02 ppm TAN as prefered range for fish culture.

Swann (no date) cited 0.0125 ppm un-ionized ammonia as upper limit for salmonids.

Kaiser et al. (1998) observed ammonia, pH and temperature peaks reaching ~0.02 mg per liter, ~8.5 and  ~26ºC or 0.0032 mg of un-ionized ammonia per liter. 

Lim et al. 2002 : "The water quality parameters that were monitored were temperature (range 26.0-27.5ºC), pH (6.4-6.6), dissolved oxygen (7.5-7.9 mg/L), ammonia (<0.02 mg/L) and nitrite (<0.1 mg/L), and all were found to be within optimal ranges for the fish tested, with no significant differences in these water quality parameters among the aquaria." It means an un-ionized ammonia level of 0.001mg per l.

Nitrite Toxicity.

The bacterias which transform nitrite to nitrate works and grow faster than the ones responsible for the conversion of ammonia to nitrite, but nitrite consuming bacterias absolutely need nitrite produced by ammonia consuming bacterias,  so unless at filter start-up phase  or under trouble conditions with biofilm usually nitrite is no problem.

 Hochheimer and Wheaton (no date) cited  0.01 mg nitrite nitrogen per liter  as a requirement of striped bass.

Buttner et al. (no date) mentioned less than 1 ppm nitrite as prefered range for fish culture but only 0.1 ppm in soft water.

Swann (no date) cited 0.1 ppm in soft water and 0.2 ppm nitrite in hard water as upper limit for salmonids, or 0.03 to 0.06 ppm nitrite-nitrogen

"Generally, freshwater-inhabiting species of the Salmonidae, Ictaluridae, Cichlidae and Cyprinidae are quite susceptive to nitrite toxicosis, and nitrite-nitrogen concentrations less than 0.5 mg/l may cause acute methemoglobinemia if environmental chloride concentrations are low." Tucker (1993). 

"The susceptibility of tropical aquarium fish to nitrite is unknown; however, it is best to keep levels low ( < 0.10 mg/l) to avoid any possible toxicity. Long-term (over 6 months) exposure to even very low nitrite levels (0.015 to 0.060 mg/l nitrite-nitrogen) can result in milt methemoglobinemia in some fish (Wedemeyer & Yasutake, 1978)" (Noga, 1995).

"The amount of nitrite entering the blood depends on the ratio of nitrite to chloride in the water. Chloride levels can be increased to lessen the effects on nitrite toxicity. At least a 20:1 ratio of chloride to nitrite-nitrogen (Cl:NO2-N) is recommended for channel catfish in ponds and tilapia and rainbow trout. Chloride levels can be increased by adding ordinary salt (sodium chloride) or calcium chloride." (Ebeling, no date).

Nitrate toxicity.

"Nitrate is the end product of nitrification and is the least toxic of the nitrogen compounds, with a 96-h LC values usually exceeding 1000 mg NO3-N/l. In recirculation systems, nitrate levels are usually controlled by daily water changes. In systems with low water exchange or high hydraulic retention times, denitrification has become increasingly important." (Ebeling, no date).

Losordo and Hobbs (2000) suggested that tilapias could be raised in water containing 150 mg nitrate-nitrogen per liter.

"Nitrate, the end product of nitrification, is relatively nontoxic except at very high concentrations (over 300 ppm). Usually nitrate does not build up to these concentrations if some daily exchange (5 to 10 percent) with fresh water is part of the management routine." Masser et al. (no date). 

Swann (no date) cited 0 to 3 ppm nitrate as tolerance range for salmonids.

"Although tolerances to nitrate toxicity vary greatly (some species can tolerate an indication of 300-400 ppm or more) the concentration should be kept well below 100 ppm: 35-70 mg/L nitrate would be better for most freshwater fishes." Petsforum. But I have doubts about what they call nitrate in this piece I cited, if it is nitrate itself or nitrogen nitrate.

"In trout production, growth of fish is affected by NO3 contents exceeding 180 mg N/l (Berka et al., 1981)." Schuster and Stelz (1998).

In my case I'll assume a maximum nitrogenous waste level as 70 mg of nitrate nitrogen per liter to size the biofilter. If your idea is to perform water changes to control ammonia (system non provided by a biofilter) I suggest you carefully take into account diurnal fluctuations on pH, that would be higher at sunset or at the time when the lights are off,  before you set a target ammonia nitrogen level. In short, you would manage to avoid high un-ionized ammonia concentrations.  

 

4.3 - Maximum Total Suspended Solids.

This is the maximum concentration of suspended solids desired to be present in the water, it is expressed as milligram per liter (mg/l). 

Suspended solids removal matters because they can clog the biofilter, and their digestion increase oxygen demand and carbon dioxide production in the system since it is consists mainly of feces and uneaten feed.

 Hochheimer and Wheaton (no date) cited  total suspended solids level below 80 mg per liter as a requirement of striped bass. 

Losordo and Hobbs (2000) suggested  10 mg suspended solids per liter as desired level for tilapias.

Swann (no date) cited up to 80 mg per liter suspended solids as upper limit for continuous exposure  for salmonids.

I'll set 10 mg of suspended solids per liter as target level.

 

4.4 - Maximum Carbon Dioxide Level. 

This is the maximum concentration of carbon dioxide (CO2) desired to be present in the water, it is expressed as milligram per liter (mg/l). 

Carbon dioxide is produced by the respiration of fishes and bacterias living in the system. 

Colt (1986): when discussing about the design of flow-through systems for salmonids he stated "Carbon dioxide concentrations should be maintained less than 20 mg/l. although higher levels can be tolerated depending on the dissolved oxygen concentration and temperature." Also he assumed that all excreted carbon dioxide remains in the water as dissolved carbon dioxide, we'll assume the same will happens here for a while. 

"Free carbon dioxide concentrations in surface waters are usually less than 10 mg/l and vary diurnally; highest concentrations occur after daybreak and lowest concentrations in mid afternoon. ... Free carbon dioxide concentrations less than 10 mg/l are usually well tolerated by fish if dissolved oxygen concentrations are near saturation. ... Prolonged exposure of fish to free carbon dioxide concentrations greater than 10 to 20 mg/l has been implicated as causing mineral deposits within kidney tubules, collecting ducts and ureters." Tucker (1993).

Masser et al. (no date)  recommended less than 20 mg CO2 per liter as requirement in recirculating systems. "Fish begins to stress at carbon dioxide concentrations above 20 ppm because it interferes with oxygen uptake."

Hochheimer and Wheaton (no date) cited up to 15 mg CO2 per liter as target level for striped bass rearing.

Buttner et al. (no date) mentioned less than 10 ppm CO2 as prefered range for fish culture.

Swann (no date) cited 0 to 10 ppm carbon dioxide  as tolerance range for salmonids.

I'll set 10 mg of CO2 per liter as desired level. 

Just to remember like I'm  designing a trickling filter and I expect it do not increase carbon dioxide level in the water but promote some CO2 stripping (degassing).

4.5 - Minimum Dissolved Oxygen Level.

This is the minimum concentration of dissolved oxygen  desired to be present in the water, it is expressed as milligram per liter (mg/l). 

Back to Kramer and Mohegan (1981) study, already discussed in Part II, I saw that guppies assumed aquatic surface respiration, a behavior intended to meet oxygen demand in water with low oxygen content, at around 3 ppm dissolved oxygen. 

Buttner et al. (no date): "During the day oxygen is produced by photosynthesis, the process by which green plants convert water and carbon dioxide in the presence of light, to oxygen and carbohydrates. During the night and day oxygen is consumed by respiration, the process by which plants and animals use oxygen to produce carbon dioxide as they burn carbohydrates, but in the day photosynthesis usually produces more oxygen than is used. Typically, oxygen levels are lowest just before dawn and highest in the late afternoon. ... As a rule of thumb, DO should be maintained above 3.0 ppm (parts per million; frequently used interchangeably with milligrams per liter, mg/L) and 5.0 ppm for warm and coldwater fish, respectively." 

Masser et al. (no date): "Dissolved oxygen (DO) concentrations should be maintained above 60 percent of saturation or above 5 ppm for optimum fish growth in most warm water systems. It is also important to maintain DO concentrations in the biofilter for maximum ammonia and nitrite removal. Nitrifying bacteria become inefficient at DO concentrations below 2 ppm." 

Swann (no date) cited oxygen levels between 5 ppm and saturation as tolerance range for salmonids. 

Hochheimer and Wheaton (no date) cited above 5 mg dissolved oxygen per liter as target level for striped bass rearing.

"Chronic hypoxia does not kill fish outright but causes considerable stress. At least 5 mg/l of dissolve oxygen is needed for optimal growth and reproduction of most fish. ... Many warm water fishes can survive for long periods in 2 to 3 mg/l oxygen, while many cold water species (e.g., salmonids) only tolerate 4 to 5 mg/l oxygen for long periods." Noga, (1995).

"The critical oxygen concentration depends on species, acclimation, activity, size temperature and other environmental conditions. Generally warm-water species survive for long periods at dissolved oxygen concentrations as low as 2 or 3 mg/l.  many cold-water fishes tolerate 4 or 5 mg/l indefinitely. The United States Environmental Protection Agency (1976) sets a minimum of 5 mg/l as the criterion for maintaining a good fish population. The embryonic and larval stages of fish are particularly intolerant of low dissolved oxygen concentrations because their respiratory and circulatory systems may not be fully developed. The early live stages of fish also have a limited ability to enviroregulate by moving away from areas of low oxygen." Tucker (1993).

I'll set 5 mg/l of dissolved oxygen as target concentration.

 

4.6 - Oxygen Saturation Level.

This is the concentration of dissolved oxygen that your water would hold if in gaseous equilibrium state with air, it is expressed as milligram per liter (mg/l). 

"The solubility of oxygen in water decreases as water temperature and salinity increase. Barometric and hydrostatic pressure also influence the solubility of oxygen. The major factor influencing barometric pressure is elevation above sea level." Tucker (1993).

An oxygen saturation calculator is available from Aquanic. 

Well, assuming a 0.2% salinity or 2 g/l or 2 ppt like recommended by Masser et al. (no date), assuming a maximum temperature of 35ºC and 0 elevation, I see my water would hold 6.876 mg of dissolved oxygen per liter. Like I expect have no significant phytoplacton community in my system that both produce and consume oxygen from the water and like I expect that the movement of water through the trickling filter and being squirt into the tanks increase oxygen levels in the water I'll assume for now that water will enter the aquariums saturated of oxygen. Anyway this supposition demands a few more studies to be confirmed.

4.7 - Water Change Required to Keep Nitrogenous Waste Level.

This is the volume of water to be changed at the given intervals (item 4.1) to keep maximum concentration of  nitrogenous waste below a given level (item 4.2), it is expressed as liters (l) in the left column and as percentage of the total volume of water in rearing units.

Keeping all assumptions I already set my model estimated that I should perform  10.10% daily water changes to control nitrate.  

4.8 - Water Change Required to Keep Total Suspended Solids Level. 

This is the volume of water to be changed at the given intervals (item 4.1) to keep maximum concentration of  total suspended solids below a given level (item 4.3), it is expressed as liters (l) in the left column and as percentage of the total volume of water in rearing units.

Here we arrived to a crossroad, my model suggested a 600% daily water change, not feasible at all. Suspended solids, fecal matter, particulate or whatever you want to call this stuff is quite dangerous for biofilters (Zhu and Chen, 2001). Since I has been breeding guppies for some time without any filter and without any problem that I could link to excessively high amounts of suspended solids in the water, and since I plan set a filter just before biofilter to avoid particulate matter clog it and to capture matter which could support the growth of non-nitrifying bacterias in the biofilter I propose don't take care of this water parameter wasting water.     

4.9 - Water Change Required to Keep Carbon Dioxide Level.

This is the volume of water to be changed at the given intervals (item 4.1) to keep maximum concentration of  carbon dioxide below a given level (item 4.4), it is expressed as liters (l) in the left column and as percentage of the total volume of water in rearing units. Just to remember like I'm  designing a trickling filter and I expect it do not increase carbon dioxide level in the water but promote some CO2 stripping I'm not taking into account in the mass balance the amount of carbon dioxide produced by nitrification process.

Another trap, my model suggested a 560% daily water change to get rid dioxide carbon. Many sources say that since CO2 concentration in the air is much lower than CO2 concentration in the water and it is passed away easily with water movement and/or by means of a "packed column" that is nothing more than a different name for trickling filters. I had no access to articles which could allow me estimate degassing in such equipments and/or trickling filters but surely I'll cover this topic in my next piece,  Part III. For now I propose do not take it into account to estimate water changes.  

4.10 - Water Change Required to Keep Dissolved Oxygen.

This is the volume of water to be changed at the given intervals (item 4.1) to keep minimum concentration of  dissolved oxygen above  a given level (item 4.8), it is expressed as liters (l) in the left column and as percentage of the total volume of water in rearing units.

My model calculated 4708% daily water changes to keep dissolved oxygen level above 5 mg/l... No comments...

4.11 - Recommended Water Change. 

This is the minimum amount of water to be changed, every a given interval of time (item 4.1), to keep nitrogenous waste under control. It is expressed as liters (l) in the left column and as percentage of the total volume of water in rearing units.

First I must to recognize that aeration and filtration systems already being used are doing a great job allowng us to reduce waste water from theoretical thousands percent per day to less than one hundred percent per week.

My model recommendation to change ~10% of water daily  is higher  than recommended by IFGA (20-40% change per week), is the upper limit reported by  Itzkovich (2002) (5-10% change daily) and  less than reported by Fernando & Phang (1985) (2/3 change every 1 to 3 days). Despite carrying capacity I proposed be the same as suggested  by IFGA I overestimated my feeding rates. I observed that a teaspoon (5 ml) of Tetramin flakes weight around 1.5 g when crumbled between the fingers to attain a size that adult guppies can easily handle, by my estimative this amount would be enough to feed seven and half  2 g guppies per day (1.5 g feed / 0.2g feed per fish per day), sincerely it is much more than I see my friends feeding to their IFGA style fishes. Itzkovich (2002) reported feeding rates of 5% body weigh daily, if I set my estimative of feeding rate at this same level my estimative of water will fall to 5% too. Until now I has been performing  50% water changes twice a week and I had stocked between 0.25 to 0.5 fish per liter, so, at least for me  and until this point, a biofilter would be water and labor saving, but I see that these savings could be much improved with better feeding practices, taking into account the nitrification occurring in aquarium walls and the increasingly resistance of fishes to nitrogenous waste as they become used to it. 

Free coppies of this spreadsheet are available for free, just drop me an email

Further References. 

Arthur, J. W., C. W. West, K. N. Allen and S. F. Redtke. 1987. Seasonal toxicity of ammonia to three fish and nine invertebrate species. Bull. Environ. Cont. Toxicol. 38:324-331.

Berka, R., Kujal, B., Lavicky, J., 1981. Recirculating systems in Eastern Europe. Proc. World Symp. on Aquaculture in Heated Effluents and Recirculation Systems, vol. 2, Stavanger, 28–30 May 1980, Berlin.

Ebeling, J. M. No date. Biofiltration. AES Workshop: Intensive Fin-Fish Systems and Technologies. p 47-56.

Hochheimer, J. N. and F. W. Wheaton. No date. Biological filters: Trickling and RBC design. p 291-318.

Itzkovich, J. 2002. Guppy culture thrives in Israel. Infofish International 2002(4):45-47.

Kramer, D. L. and J. P. Mehegan. 1981. Aquatic surface respiration, an adptative response to hypoxia in the guppy, Poecilia reticulata (Pisces, Poeciliidae). Env. Biol. Fish. 6(3-4):299-313.

Meade, J. W. 1985. Allowable ammonia for fish culture. Progressive Fish Culturist 47:135-145.

Noga, E. J. 1995. Fish Disease: diagnosis and treatment. Mosby-Year Book, Inc. St. Louis.

Smith, C. E. and R. G. Piper. 1975. Lesions associated with chronic exposure to ammonia. In: Ribelin, W. E. and G. Migaki. The Pathology of Fishes. Univ. Wisconsin Press. Madison. p 497-514.

Tabata, K. 1962. Toxicity of Ammonia to Aquatic Animals with Reference to the Effect of pH and Carbonic Acid. (English translation used.) Tokai-ku Suisan Kenkyusho Kenkyu Hokoku 34:67-74.

Tucker, C. S. 1993. Chapter 13 - Water Analysis. in Stoskopf, M. K. (ed). Fish Medicine. W. B. Saunders Co. Philadelphia. p 166-197. 

U. S. EPA. 1976. Quality Criteria for Water. EPA-440/9-76-023. United States Environmental Protection Agency. Washington, DC.

Wedemeyer, G. A. and W. T. Yasutake.  1978. Prevention and treatment of nitrite toxicity in juvenile steelhead trout (Salmo gairdneri). J. Fish. Res. Bd. Can. 35:822-827.