RAS-FARM DESIGN SHOULD BE BASED ON DATA

RAS-farm design should be based on data. RAS-farms should be designed with data in mind to gain optimal performance, Blue Unit suggests. It can save both time and resources.

The modern RAS-farm is complex and is made up of large and costly machines, such as fish tanks, drum filters, biofilters and degassers. If a component from a RAS farm does not operate correctly, it will have a negative impact on the entirety of the farm’s functions. Today, there are several methods and machines to measure output of RAS-farms for optimal performance, but these are generally believed to be linked with uncertainty and lack of holistic insight into the farm’s function.

Blue Unit works on the basis that future decision-making will be based on data, and data collection will therefore have a much greater influence on the design and operational capability of RAS-farms. This will demand more widespread application of centralised sensor-solutions and improved alarm systems based on gradients rather than raw data.

Data is the future

In collaboration with several international clients, Blue Unit has worked on, evaluated and re-evaluated and prepared new and better ways to utilise data with a single purpose: optimising RAS-farm performance.

“In the RAS-business, there’s a tendency to design farms out of assumptions gained from practical experience rather than raw data. For example, degassers are usually dimensioned according to carbon dioxide removal rate. But it’s naive to presume that carbon dioxide alone influences the function of a degasser. There is a myriad of parameters at stake,”  says David Owen, CTO of Blue Unit.

Blue Unit predicts the RAS-industry will soon begin implementing intelligent farm control, where operational data is calculated and sent back to the operating systems of local farms, so the machines can adjust in real time. However, this requires high quality data to succeed, which can for instance be gained through centralised sensor networks.

Machines should measure machines

By measuring water quality from either side of the various components on the RAS-farm, important water quality parameters, such as CO₂-levels and turbidity, are measured. Reliable data is everything, if measurements are to be interpreted correctly by monitoring precise water quality gradients.

Today, most RAS-farms utilise decentralised sensors – sensors placed at various locations on the farm. This could for instance be a pH-sensor in the main inflow to the fish tank. Hanging sensors on either side of the farm’s machines creates data that is hard to compare due to the sensor’s uncertainty, such as operational fluctuations or different levels of cleanliness. These errors and fluctuations give an uncertain basis for calculating water quality gradients.

“For the last 10 years, we at Blue Unit have focused on developing centralised water quality monitoring systems and continuously work towards improved data implementation in the new generation of data-fuelled fish farming,” says David Owen.

Water quality gradients can be used for further analyses, where performance and potential issues can be identified. This could for instance be early signs of bacterial activity in a degasser, leading to lowered CO₂-emission in the water. These insights demand high quality data along with an understanding of normal operations, which is also the reason for the use of machine learning and artificial intelligence to create smarter farms.

According to Blue Unit, centralised sensors are the most practical and safe way to monitor water quality gradients. By utilising centralised sensor-systems to monitor water quality, measurements can be taken from both in- and outflow of the fish tank, of the mechanical filters, biofilters and degassers. This gives a deeper insight into how the farm functions in real time while keeping uncertainty in measurements to a minimum.

Facts: Blue Unit Lab Station

• All-in-one water quality monitoring system with an option for Cloud-based software that collects data and gives 24/7 farm status insight
• Specially designed and industrially produced sensors of high quality
• Monitors: pH, Oxygen, opaqueness, conductivity, redox, rH (redox lesser pH), salt contents, temperature, total dissolved matter, available CO₂, total CO₂, H₂S, non-carbonate alkalinity (surrogate of ammoniac)
• Unique insight into farm performance on a global level
• Up to 2700 datapoints daily that are uploaded to the Cloud and can be accessed from portable devices
• Early Warning that shows where, when, and why changes occur in water quality
• Built-in alarm system at critical levels
• Compares benchmark-values with similar businesses in the industry