## Dismantling the "Equipment Stacking" Myth: The Foundational Logic of RAS from a Systems Engineering Perspective Over the past few years, during the deployment of actual projects across Singapore and China, our team has participated in dozens of technical review meetings for large-scale Recirculating Aquaculture Systems (RAS). The focal point of these discussions is often surprisingly uniform: "How powerful a water pump do we need? Should we purchase micro-screen drum filters from Europe, the US, or China?" In reality, this trajectory is misguided. Regardless of how expensive the equipment is, significant outcomes cannot be achieved through sheer brute force. We must prioritize a fundamental calculus: the mass balance of the system. Assume a system feeds 1,000 kilograms of feed into aquaculture tanks daily. The protein content of fish feed typically ranges from 30% to 55%. Based on this, approximately 3% to 5% of the total feed weight will be directly metabolized and converted into highly toxic total ammonia nitrogen (TAN, primarily comprising ionized $NH_4^+$ and un-ionized $NH_3$) (Losordo, 2015). If this hydrological balance is not accurately calculated, even the most sophisticated hardware stack will inevitably collapse under the shock of biochemical loading. Consider the high-density commercial farms designed by established North American RAS agencies (such as PR Aqua). Why do they exhibit such notable stability? Their core philosophy is singular: treating the entire aquaculture facility as a precise, dynamic mass balance model. The "carrying capacity" of the system is meticulously derived by strictly calculating constants such as feed input, biological conversion rates, and excretion rates. Industry standards indicate that approximately 4% of the feed input will irreversibly convert into TAN within the RAS (Losordo, 2015). What does this imply? The true focus of the design is constructing a dynamic circulation network capable of precisely assimilating and converting this 4% toxin load. Carbon, nitrogen, phosphorus, and oxygen are all indispensable. Through such rigorous systems engineering, RAS can achieve exceptionally high resource retention efficiencies. For instance, in the highly intensive recirculating aquaculture case of Taste of BC Aquafarms in Canada, they ultimately achieved a total phosphorus recovery efficiency of up to 83% (Vinci et al., 2013). This is not merely environmental compliance; it is robust evidence of meticulously controlled material fluxes. Balance is inherently fragile. To maintain it, the interplay between hydraulic retention time and circulation rates becomes the lifeblood of piping network design. How does the water circulate? At what velocity? These factors directly dictate the self-purification capacity of the tanks and oxygen delivery. When constructing models, engineers must tailor parameters to the specific species: for high-oxygen-demand cold-water fish like Atlantic salmon, the water turnover rate is typically set at 30 minutes or less; conversely, for warm-water species like tilapia, a 60-minute turnover rate is often sufficient (Losordo, 2015). More critically, there is the challenge of ammonia nitrogen. To strictly suppress TAN below safe thresholds, the design flow must continuously flush through the biofilter at a high frequency of 1 to 2 times per hour (Losordo, 2015). Under highly intensive standards, the massive water volume of the entire system undergoes complete physical and biochemical renewal every 3 to 4 days (Vinci et al., 2013). In other words, for every kilogram of feed consumed on average, the system requires the intensive circulation and distribution of up to 540 liters of new water (Vinci et al., 2013). This may sound ideal, but calculating the electricity costs reveals a different reality. If such high-frequency dispatch relies entirely on the brute force of high-pressure water pumps, the energy expenses could render large-scale commercial RAS economically unviable. This constraint has necessitated another cornerstone of modern engineering design: low-head and gravity-fed layouts. By ingeniously utilizing minor elevation differences, the water flow naturally cascades between the drum filter, degassing column, and bioreactor without the need for electrical pumping. More importantly, this approach avoids the excessive shear forces exerted by high-pressure pump impellers on feces and uneaten feed. Once fecal matter is pulverized, the interception rate of subsequent physical filtration becomes largely ineffective. Naturally, there are trade-offs. Gravity flow necessitates complex, tiered facility designs, which significantly increases initial civil engineering costs. Finally, the maturity of a system is evidenced by its modular parallelization and redundancy.
