Abstract: Biological filtration in the broadest sense includes any filtration technique that utilizes biological (living) organisms to remove impurities from the wastewater. Filter media selection is critical in the operation to achieve effluent quality requirements. The most important is to choose the correct types of filter media. Laboratory studies were conducted to evaluate the optimum ammonia removal performance using four different types of filter media (Ceramic Ring A, Ceramic Ring B, Japanese Filter Mat and Filter Wool) at different ammonia loading rates of 20 until 120 mg L-1. Ceramic Ring A has been found to give the best performance with respect to their efficiency of ammonia removal because of high surface area and characteristic roughness. In general, nitrification is most efficient at pH levels ranging from about 7.5 to 9.0. Water temperature was kept between (27 and 30°C). Nitrification efficiency is slower at lower temperatures.
INTRODUCTION
Ammonia removal is accomplished through biological nitrification in which ammonia is converted into nitrite and then nitrate by nitrifying bacteria called nitrifiers. The removal of ammonia is one of the most important factors in fish culture system. In water, NH4+ (ammonium) and NH3 (ammonia) are in equilibrium depending on the pH and the temperature (Timmons et al., 2002). The sum of the two forms is called Total Ammonium Nitrogen (TAN). Ammonia is the major excretory product harmful to fish.
Biological filters are commonly used for purifying culture water in recirculating fish culture systems. Four different types of filter media are used for nitrification. The wide variety of biological filter materials available on the market today can confuse almost anyone trying to find the right one. Every manufacturer claims their filter material is the most efficient and easiest to use. In recirculating systems, biological flters are flled with media that provide the suitable environment for the nitrifying autotrophic bacteria to grow. Several natural and artificial non-toxic flter media are commonly used. For optimum nitrification rate and less clogging of the flter, flter media should have high specific surface area for bacterial adhesion, low specific gravity and clogging properties (Hagrove et al., 1996). The influence of temperature over nitrification processes is important for biological filter operation.
In a suspended culture, biological reaction rates increase with rising temperature until an optimal temperature is reached (Saywer et al., 1994). The interactions of pH, nitrification and water quality can be quite complex. In general, nitrification is most efficient at pH levels ranging from about 7.5 (Korner et al., 2001). At the higher pH ranges (8.5-9.0), nitrification rates are fastest given sufficient ammonia. However, at the low ammonia concentrations usually found in aquacultural systems, operating at a pH of about 7.0 can be efficient (Painter, 1970). The purpose of this study is to examine the effects of the temperature and pH on the ammonia removal performance in laboratory scale fish tank unit using four different types of filter media used in biological treatment system. It should be noted that for several of the flter media types used, there is currently very little knowledge available of their use in aquaculture field.
MATERIALS AND METHODS
Experimental systems: Laboratory scale fish tank unit were made of acrylic and was setup by replicating and modifying the freshwater fish water tank treatment system in Aquamarine Biological Research Centre, Penang by scaling down the sizes, amount of packing and other parameter that mainly affects the treatment operation. It consisted of a 0.56x0.45x0.17 m rectangular recycling tank with water volume was 0.025 m3 a pump, physical filtration column, a primary filtration holding tank (receiver tank), biological treatment tank and air pumps as well as other equipments that might be required for analyses the quality of water samples (Fig. 1).
| Table 1: | Characteristics of the filter media packed in the biological system |
| Fig. 1: | Laboratory scale of fish tank unit |
The study was conducted in two stages. In the first stage, experiments were conducted to observe and investigate nitrification performance efficiency among four types of different filter media characteristics for different ammonia loading rates. In the second stage, the most efficient filter media which showed excellent performance in nitrification process was then selected to further study on the ammonia removal efficiency of the biological treatment unit. The most significant factors such as temperature, Dissolved Oxygen (DO), pH and alkalinity are studied during 40 days of experiment. During the test, the hydraulic loading rate for the entire experiment was fixed at 0.32 m3 h-1 with an air flow rate of 0.2 m3 h-1 as described from the treatment unit at Aquamarine Biological Centre. The filter media tested was acclimated by increasing ammonia loading starting 20 mg L-1 with weekly increments 20 mg L-1 until 140 mg L-1 on the final week of the experiment. The medium’s performance was evaluated based on effectiveness of ammonia removal as well as overall conversion to nitrate. Collected samples were measured for ammonia, nitrite and nitrate concentration. Parameters such pH, temperature and alkalinity were measured before sampling while the temperature was monitored daily throughout the course of the experiment.
The experiment was carried out over the seven weeks period. Four different filters media, namely, Ceramic Ring A (CRA), Ceramic Ring B (CRB), Japanese Mat (JM) and Filter Wool (FW) were tested in the biological treatment system. The characteristics of these four types of media are summarized in Table 1. In this study, the effect of temperature, alkalinity and pH were only been focused and discussed.
Analytical techniques: In accordance with Standard Methods, a Hach (Model DR 2800) spectrophotometer was used to measure the different nitrogen forms colorimetrically. Direct colorimetric methods were employed to determine the concentrations of ammonia, nitrite and nitrate respectively, in the liquid medium (APHA, 2007). Measurements of temperature, DO and pH were also performed following the procedures specified in Standard Methods (APHA, 2007).
RESULTS AND DISCUSSION
In the first stage of study, throughout the experiment, DO was maintained above 6 mg L-1. The average concentrations of DO in this study were high for all flter media. These levels were apparently adequate for the colonization of the bacteria and efficient nitrification performance. It also decreased the toxicity of the encountered occasional high concentrations of ammonia and nitrites (Camargo et al., 2005). Water temperature was kept between 27 and 30°C, which are within the acceptable range for the growth of both nitrifying bacteria (Srna and Baggaley, 1975).
Effect of filter media surface area: It can be observed and summarized that for higher concentrations of ammonia, the nitrification reactions converts about 89.0-92.0% of the ammonia into nitrate compound using Ceramic Ring A while Japanese Mat showed about 76.0-86.0% conversion at higher ammonia loading. At lower concentrations, Ceramic Ring A worked very effectively, removed almost 92.0-94.0% of ammonia while the Japanese Mat showed acceptable conversion of 86.0-88.0%. Both of Filter Wool and Ceramic Ring B showed poor ammonia removal from the water at low and higher loading of ammonia (Fig. 2).
The four different types of filter media were efficient in oxidizing ammonia and in maintaining other quality parameters. When ammonia samples with different concentration (range 20-140 mg L-1) were tested, all mediums can apparently remove ammonia level to a desired level. The concentration decreased in the outlet water from all the media and in the same pattern with no much significant different level.
| Fig. 2: | Efficiency of filter media in nitrification process |
The reason it is important to consider what filter media to use when we build a biological filter systems is because different media have different amounts of surface area. The more surface area the more bacteria can colonize and live in filter media to process all of the waste or uneaten food. The surface of the media provides the support medium on which the biofilm develops. It follows that the greater the surface available for biofilm growth the greater the potential for ammonia removal. The specific surface area available for biofilm growth varies significantly between the different types of media (Table 1). In nitrifying biological filters, (Kikuchi et al., 1994) found that effective ammonia removal was related to the specific surface area of the medium. They concluded that the surface texture of media had an effect on nitrification. It has been found that the higher the available surface area of biological filter medium, the greater the rate of nitrification per volume of media, provided that the interstices between the medium did not block with solids. The degree of blocking of interstices is related to media size, voidage and loading rates. The work performed (Balakrishnan and Eckenfelder, 1969), also found that increasing the specific surface area of plastic media from 190 to 250 m2 m-3 resulted in an increase of the nitrification efficiency.
Effect of temperature: During the early stage of experiment, the temperature range is about 31 to 32°C while during the last day of experiments, it decreased until 28.5°C. It can be seen in Fig. 3, the efficiency of ammonia removal was above 93.5% during high temperatures. During lower temperature, the ammonia removal efficiency is decreased steadily as the temperature declined. However, the impact of temperature change on nitrification process was less significant because only a little change was observed on ammonia removal efficiency. It has been proved by other researcher (Zhu and Chen, 2002) studied the impact of temperature on nitrification rate through laboratory experiments, mathematical modelling and sensitivity analysis. A lower nitrification rate was observed only at the lowest temperature they tested, 8°C.
| Fig. 3: | Effect of temperature on ammonia removal efficiency in ceramic ring A filter media |
It has been well accepted that a higher temperature enhances nitrification rate as the biochemical driven bacterial processes accelerate as temperature increases (Zhu and Chen, 2002). Nitrifying bacteria are known to be very sensitive to low temperatures and nitrification can be strongly inhibited by low temperatures. Low temperatures are known to limit the activity of the microorganisms and very low temperatures can result in deactivation. High temperatures can cause the death of the microorganisms (Jones and Morita, 1985).
Effect of pH and alkalinity: For the experiment conducted, alkalinity is naturally decreased from time to time (Fig. 4). The amount of alkalinity decreased in this study is from 252 mg L-1 to 135 mg L-1 for the first 5 days of the experiment while for the pH is declined from 8.2 to 7.8. The reduction of alkalinity in filter medium is caused by nitrification process during the acclimation period. On day 7 of experiment, the alkalinity dropped below 150 mg L-1 CaCO3 which is below than the requirement for the nitrification process as well as pH maintenance. Sodium bicarbonate (NaHCO3) is added to reduce the toxic effects of the nitrite and this also helps against the ammonia. It is considered the safest and easiest buffer to add as it dissolves rapidly and is much less dangerous to handle. Routine alkalinity adjustments were required because the nitrifying bacteria continually consume alkalinity at a rate of about 7 mg (expressed as CaCO3) per milligram of total ammonia nitrogen oxidized to nitrate. As shown in Fig. 3, the initial pH value was maintained at desirable range (6.5 -8.0). Chen et al. (1989) showed that the rate of nitrification would be reduced when alkalinity was below 40 g m-3. (Gujer and Boller, 1986) reported that in nitrifying bioflters used in municipal wastewater treatment, an alkalinity level of at least 75 mg L-1 (g m, o r 1.5 meq L-1) was needed to maintain maximum nitrification rate.
A great deal of investigation conducted has demonstrated the pH effects on nitrification. Based on the review provided by other researchers’ studies (Zhu and Chen, 2002), the optimal pH for the growth of nitrifying bacteria varies widely.
| Fig. 4: | Relationship in pH and alkalinity in the biological treatment unit in ceramic ring A filter media |
The optimum pH for nitrification can range from 7.0 t o 9.0, while the optimum pH range from 7.2 to 8.8 for Nitrosomonas and 7.2 to 9.0 for Nitrobacter. Therefore, reduced nitrification activity at lower pH levels may result indirectly from substrate limitation since the fraction of NH3–N in total ammonia nitrogen decreases with the decrease of pH (Allison and Prosser, 1993). According (Villaverde et al., 1997) the optimal range of pH for nitrification can be determined by the three different effects that the pH can exert on nitrifying bacteria: (1) activation deactivation of nitrifying bacteria; (2) nutritional effect, connected with alkalinity; (3) inhibition through free ammonia and free nitrous acid. Flora et al. (1999) mentioned that nitrification rates may be improved by increasing bulk pH to high (alkaline) values. This is very significant for the operation of nitrification bioflters in aquaculture systems.
CONCLUSIONS
This study evaluated the efficacy of the biological process treatment of the aquaculture using different types of flter media. The following conclusions can be drawn:
The highest ammonia removal in this study was obtained in the columns filled with Ceramic Ring A, the largest specific surface area. This is expected, since a larger surface will lead to a larger area of biofilm.
Although a low pH is optimal in the culture tank to minimize the toxicity of the portion of unionized ammonia, regardless of the optimal pH for the growth of aquaculture species, a pH higher than the optimum in the biological treatment system is desirable for the improvement of the nitrification efficiency.
The results of this study demonstrated that the temperature impacts on the ammonia removal were not as significant as predicted.
ACKNOWLEDGMENT
The authors gratefully acknowledge the research funding provided by USM short term grant and USM fellowship.