8.2 Mineralization Loop
In RAS, solid and nutrient-rich sludge must be removed from the system to maintain water quality. By adding an additional sludge recycling loop, accumulating RAS wastes can be converted into dissolved nutrients for reuse by plants rather than discarded (Emerenciano et al. 2017). Within bioreactors, microorganisms can break down this sludge into bioavailable nutrients, which can subsequently be delivered to plants (Delaide et al. 2018; Goddek et al. 2018; Monsees et al. 2017a, b). Many one-loop aquaponics systems already include aerobic (Rakocy et al. 2004) and anaerobic (Yogev et al. 2016) digesters to transform nutrients that are trapped in the fish sludge and make them bioavailable for plants. However, integrating such a system into a coupled one-loop aquaponics system has several disadvantages:
The dilution factor for nutrient-rich effluents is much higher when discharging them to a single-loop system in relation to discharging them to the hydroponics unit only. Effectively, nutrients diluted by entering in contact with large volumes of fish rearing water.
Fish are unnecessarily exposed to the mineralization reactor’s effluents; e.g. the effluents of anaerobic reactors can include volatile fatty acids (VFAs) and ammonia that might potentially harm the fish; such reactors also represent an additional source for potential introduction of pathogens.
Around 90% of the nutrients trapped in the sludge can be recovered when RASsludge is maintained at a pH of 4 (Jung and Lovitt 2011). Such a low pH is not possible when operating bioreactors at a pH around 7 (Goddek et al. 2018), which is the usual trade-off pH value within one-loop aquaponics systems.
Fig. 8.3 Approximate pH of the water within the different system components as well as the process water. The ‘~’ indicated an approximation
With respect to pH, Fig. 8.3 shows the approximate pH values of the respective process water flows in a multi-loop aquaponics system (e.g. as presented in Fig. 8.1c). Figure 8.3 also shows the impact of mineralization reactors on the performance of the system as a whole, based on the anaerobic reactors proposed by Goddek et al. (2018). Such a system represents only one possible solution for treating sludge, with alternative approaches discussed in Chap. 10. The decrease in pH of the process water flowing from the RAS subsystem into the hydroponics loop as shown in Fig. 8.3 demonstrates acidification in the nutrient concentration loop (i.e. demineralized water has a pH of 7). Thus, the effluent has a lower pH than the RAS outlet, which reduces the need to adjust the pH for optimal plant growth conditions.
Table 8.1 Overview of optimal growth conditions for fish and plants and preferred operational conditions for sludge nutrient recycling treatment
table thead tr class=“header” thSubsystem/th thSpecies/ function/th thpH/th thTemperature (˚C)/th thNitrate (NOsub3/sub) (mg/L)/th /tr /thead tbody tr class=“odd” td rowspan=“2"Recirculating aquaculture system (RAS)/td tdiOreochromis niloticus/i (Nile tilapia)/td td7–9 (Ross 2000)/td td27–30 (El-Sayed 2006)/td td<100–200 (Dalsgaard et al. 2013)/td /tr tr class=“even” tdiOncorhynchus mykiss/i (rainbow trout)/td td6.5–8.5 (FAO 2005)/td td15 (Coghlan and Ringler 2005)/td td<40 (Davidson et al. 2011; Schrader et al. 2013)/td /tr tr class=“odd” td rowspan=“2"Hydroponics/td tdiLactuca sativa/i (lettuce)/td td5.5–6.5 (Resh 2012)/td td21–25 (Resh 2012)/td td730 (Resh 2012)/td /tr tr class=“even” tdLycopersicon esculentum (tomato)/td td6.3–6.5 (Resh 2002)/td td18–24 (Resh 2002)/td td666 (Sonneveld and Voogt 2009)/td /tr tr class=“odd” td rowspan=“2"Anaerobic reactor/td tdMethanogenesis/td td6.8–7-4 (de Lemos Chernicharo 2007)/td td30–35 (Alvarez and Lidén 2008; de Lemos Chernicharo 2007)/td td–/td /tr tr class=“even” tdSludge mobilization/td td4.0 (Jung and Lovitt 2011)/td tdn/a/td td–/td /tr /tbody /table The two-stage reactor system works as follows:
In the first stage (pH around 7 to provide optimal conditions for methanogenesis; Table 8.1), the organic matter is broken down to sustain a high degree of methane production (i.e. carbon removal). Mirzoyan and Gross (2013) reported a total suspended solids reduction of around 90%, using upflow anaerobic sludge blanket reactor technology. This has the benefit that (1) biogas is harvested as a renewable energy source and (2) fewer VFAs are produced in the second stage. The sludge retention time in the first stage should be several months, before removing the accumulated nutrients in the sludge (e.g. calcium phosphate aggregation) within the second stage.
In the second stage, nutrients in suspended solids are effectively mobilized and become available for plant uptake. This mobilization is the most effective in a low-pH environment (Goddek et al. 2018; Jung and Lovitt 2011). Once the pH of acidic reactors is decreased, it usually remains stable; thus less pH regulation is required in the hydroponic unit.
The effluents that are rich in nutrients may require some post-treatment depending on the amount of measured total suspended solids and VFAs. However, it is important to keep in mind that ammonia can stimulate plant growth, e.g. leafy greens, when it accounts for 5—25% of the total nitrogen concentration (Jones 2005). However, fruit vegetables such as tomatoes or sweet peppers are particularly sensitive to ammonia in the nutrient solution. An aerobic post-effluent treatment or a well-aerated hydroponics sump would be required in systems growing those types of crops.
8.2.1 Determining Water and Nutrient Flows
For system sizing (Sect. 8.4), the amount of water flowing from the RAS system via the reactor(s) to the hydroponics unit (QsubMIN/sub) needs to be known (Eq. 8.1):
$Q_{MIN} (kg/day) = \frac{n_{feed} \times k_{sludge}}{ \pi_{sludge}}$ (8.1)
where nsubfeed/sub is the amount of fish feed in kg, ksubsludge/sub is the proportion coefficient of fish feed ending up as sludge, and πsubsludge/sub is the proportion of total solids (i.e. sludge) in the sludge water flow entering the mineralization loop.
The sludge concentration can be increased by adding a gravity separation device prior to the bioreactors, directing the ‘clear’ supernatant back to the RAS system. This formula can also be used to get an input for sizing the reactor based on the hydraulic retention time (Chap. 10). Between 20 and 40% of the fish feed ends up as total suspended solids in the RAS-derived sludge (Timmons and Ebeling 2013). As an example, it has been found that tilapia sludge contains around 55% of nutrients that were added to the system via feed (Neto and Ostrensky 2013; Yavuzcan Yildiz et al. 2017) which represents a valuable resource for crop growth.
The main nutrients that can be recovered via a mineralization process are N and P. As P (one of the major components of sludge) is the most valuable macronutrient in terms of cost and availability for crop production, it should be the first element to be optimized in the aquaponic system.
The mineralization rate of the mineralization loop is calculated as follows:
$Mineralization (g/day) = (n_{feed} \times 1000)π_{feed}\times π_{sludge} \times η_{min}$ (8.2)
where _n_subfeed/sub is the feed input to the system (in kg); _π_subfeed/subis the proportion of the nutrient in the feed formulation;_π_subsludge/subis the proportion of a specific feed-derived element ending up in the sludge; and ηsubmin/subis the mineralization and mobilization efficiency of the reactor system.
The last step would be to determine the concentration of the respective element in the effluent of the mineralization loop:
$Nutrient\ concentration\ (mg/L) = \frac{Mineralization \times 1000}{Q_{MIN}}$ (8.3)
Example 8.1
Our RAS system is fed with 10 kg of fish feed per day. We assume that 25% of the fed feed ends up as sludge. In our system, we use a Radial Flow Settler (RFS) to concentrate the sludge to 1% dry matter. Consequently, the flow from the RAS to HP via the mineralization loop is calculated as follows:
$Q_{MIN}\ (kg/day) = \frac{10kg\ \times\ 0.25}{0.01}=250\approx 250kg/day$
We decide to size our system on P. The P content of our feed (in most cases provided by the feed manufacturer) is 1.5% and 55% of it ends up in the sludge (Neto and Ostrensky 2013). We assume that our reactors achieve a mineralization efficiency of 90% for this element. Therefore, the grams of P transferred to the hydroponics unit each day can be determined:
$Mineralization\ (g/day)=(10kg\times 1000)\times 0.55\times 0.015 \times 0.9=74.25$
The concentration of the effluent in consequently:
$Nutrient\ concentration\ (mg/L)=\frac{74,25g \times 1000}{250L}=297\ mg/L$
This concentration of P in the effluent in the example box above is approximately six times higher than in most hydroponics nutrient solutions. The research of Goddek et al. (2018) underpins this theoretical number, and they report that their RAS sludge contained 150 and 200 mg/L of P for two independent systems, respectively (1% TSS sludge), with a fish feed P content of 0.83% in dry matter feed for the latter (200 mg/L).