Design Flow Rates for Commercial Aquaponics

UVI aquaponics design flow rates

Now let’s get a little more into the design of an aquaponic system at a commercial scale.  We will use the UVI aquaponic system again for this example because it has measurable data available.  The previous article showed some of the effects protein and sunlight have on the design parameters of commercial aquaponics.  In this article I want to discuss the hydraulics and design flow rates and how simple changes in just protein and to a lesser degree temperature will change your design very quickly.UVI aquaponics design flow rates

The above image is a basic layout of one of the UVI aquaponic systems.  There are some minor anomalies in the water volumes (missing 1400 litres which we have taken from the filtration to allow for free board).  the hydroponic subsystem, depending on which publication you read changes in volume.  So will stick with one and go with that.

To explain briefly how the table/drawing works the following will give you some insight.  The black numbers represent the volume of water in each section, the red numbers represent the time in minutes to exchange the water through that section and is governed by the pump rate.  Rakocy et al published an increase from 163lpm to 378lmp was required to maintain water quality.  I have set all the other parameters (excluding oxygen) to meet that flow rate.  So we can see how the design flow rate is one of the most important aspects in your planning.  And perhaps that is why I go on about it so much…

The table as it stands is set to most of the information from UVI.  I will draw your attention to the oxygen in the top table which is highlighted and the design flow rate (868lpm) which is more than double that of the design flow rate of 378lpm.  We are only allowing for the bacterial oxygen demand in that table as the tanks and rafts are aerated to maintain an outlet from the raft beds of 6.9mg/L of dissolved oxygen before it is pumped back to the culture tanks.  But….  that was only for the first 12 weeks of the 24 week production cycle.  The final 12 weeks required an additional vertical lift pump to maintain oxygen as the biomass increased.  Perhaps if the flow rate was increased to the design in this table, that may not have been required.

I have made some pretty large assumptions in the efficiency of each of the treatment units setting all but the oxygen to 100%.  If we were to reduce their efficiency each treatment unit would require a higher pumping rate as shown in the table below where we have set the efficiency to 70%.  It makes sence that if you are trying to maintain a minimum or maximum level of wastes and the treatment does not achieve that in one pass (100%) then the flow rate must increase respectively.

Aquaponic treatment efficiencyYou can see some immediate issues with the 378lmp design flow rate in the solids filtration.  The UVI aquaponic system uses clarifiers that generally require at least 20 minutes retention time where as at this flow rate they are running much higher at 9.5 minutes.  The less solids captured in the clarified increased mineralization and an increase in nutrient in the system however the bio filter (netting) required cleaning more often.  The bio-filtration is just a little on the small side as well with a retention time of just 7.4 minutes so there will be some reliance on the hydroponic system surface area to do some ammonia processing.  Not a huge concern because of the size of the raft system.

Now those results are based on a maximum temperature of 29c and a feed protein content of 32% which is suitable for Tilapia.  The image/table below shows what changes if we keep the same temperature but change our fish species to say Barramundi which requires between 45% and 50% protein in their feed.

Barramundi in aquaponics

In the above table we have set the design flow rate for Nitrate as we have with the first Tilapia table.  Note now with an increase in protein from 32% to 45% (29% increase) our design flow rate (pump rate) is now 532 litres per minute which is also a 29% increase.  Also both the solids and biological filtration will not be that effective with those limited retention times.

What is of particular interest is the hydroponic system has now increased in volume and in area, 214 m2 to 301 m2 and 11.5 m3 to 16.2 m3 respectively.  Both increased reflect the same in crease in flow rate and protein increase.  This confirms the primary design criteria is the input of the feed and the protein content in the feed.   During the increases we have managed to decrease the retention time on the fish tanks which now exchange nearly 2 times per hour but we maintained the 3 hour exchange rate on the hydroponic system.

If you were to put Rainbow Trout in that system, running at much cooler temperatures (18c) with the same protein the changes are minor but some of the water quality criteria will have to change before calculating design flow rates.  You would see a decrease in flow rate but the solids filtration would have to be improved.

I have not compared fish density and nutrient dilution at this point, that is a topic of another conversation.  We may also take a look at what happens when you add a media grow bed to the mix…

Regards
Paul

Now let’s get a little more into the design of an aquaponic system at a commercial scale.  We will use the UVI aquaponic system again for this example because it has measurable data available.  The previous article showed some of the effects protein and sunlight have on the design parameters of commercial aquaponics.  In this article I want to discuss the hydraulics and design flow rates and how simple changes in just protein and to a lesser degree temperature will change your design very quickly.UVI aquaponics design flow rates

The above image is a basic layout of one of the UVI aquaponic systems.  There are some minor anomalies in the water volumes (missing 1400 litres which we have taken from the filtration to allow for free board).  the hydroponic subsystem, depending on which publication you read changes in volume.  So will stick with one and go with that.

To explain briefly how the table/drawing works the following will give you some insight.  The black numbers represent the volume of water in each section, the red numbers represent the time in minutes to exchange the water through that section and is governed by the pump rate.  Rakocy et al published an increase from 163lpm to 378lmp was required to maintain water quality.  I have set all the other parameters (excluding oxygen) to meet that flow rate.  So we can see how the design flow rate is one of the most important aspects in your planning.  And perhaps that is why I go on about it so much…

The table as it stands is set to most of the information from UVI.  I will draw your attention to the oxygen in the top table which is highlighted and the design flow rate (868lpm) which is more than double that of the design flow rate of 378lpm.  We are only allowing for the bacterial oxygen demand in that table as the tanks and rafts are aerated to maintain an outlet from the raft beds of 6.9mg/L of dissolved oxygen before it is pumped back to the culture tanks.  But….  that was only for the first 12 weeks of the 24 week production cycle.  The final 12 weeks required an additional vertical lift pump to maintain oxygen as the biomass increased.  Perhaps if the flow rate was increased to the design in this table, that may not have been required.

I have made some pretty large assumptions in the efficiency of each of the treatment units setting all but the oxygen to 100%.  If we were to reduce their efficiency each treatment unit would require a higher pumping rate as shown in the table below where we have set the efficiency to 70%.  It makes sence that if you are trying to maintain a minimum or maximum level of wastes and the treatment does not achieve that in one pass (100%) then the flow rate must increase respectively.

Aquaponic treatment efficiencyYou can see some immediate issues with the 378lmp design flow rate in the solids filtration.  The UVI aquaponic system uses clarifiers that generally require at least 20 minutes retention time where as at this flow rate they are running much higher at 9.5 minutes.  The less solids captured in the clarified increased mineralization and an increase in nutrient in the system however the bio filter (netting) required cleaning more often.  The bio-filtration is just a little on the small side as well with a retention time of just 7.4 minutes so there will be some reliance on the hydroponic system surface area to do some ammonia processing.  Not a huge concern because of the size of the raft system.

Now those results are based on a maximum temperature of 29c and a feed protein content of 32% which is suitable for Tilapia.  The image/table below shows what changes if we keep the same temperature but change our fish species to say Barramundi which requires between 45% and 50% protein in their feed.

Barramundi in aquaponics

In the above table we have set the design flow rate for Nitrate as we have with the first Tilapia table.  Note now with an increase in protein from 32% to 45% (29% increase) our design flow rate (pump rate) is now 532 litres per minute which is also a 29% increase.  Also both the solids and biological filtration will not be that effective with those limited retention times.

What is of particular interest is the hydroponic system has now increased in volume and in area, 214 m2 to 301 m2 and 11.5 m3 to 16.2 m3 respectively.  Both increased reflect the same in crease in flow rate and protein increase.  This confirms the primary design criteria is the input of the feed and the protein content in the feed.   During the increases we have managed to decrease the retention time on the fish tanks which now exchange nearly 2 times per hour but we maintained the 3 hour exchange rate on the hydroponic system.

If you were to put Rainbow Trout in that system, running at much cooler temperatures (18c) with the same protein the changes are minor but some of the water quality criteria will have to change before calculating design flow rates.  You would see a decrease in flow rate but the solids filtration would have to be improved.

I have not compared fish density and nutrient dilution at this point, that is a topic of another conversation.  We may also take a look at what happens when you add a media grow bed to the mix…

Regards
Paul

About the author

Paul Van der Werf

Paul is the Operations Manager for a 4400m2 integrated aquaculture pilot project in the United Arab Emirates desert he designed and built. This is a commercial aquaponics pilot to evaluate integrated farming in arid climates.

Related posts

10 Comments on “Design Flow Rates for Commercial Aquaponics

  1. Chris

    Great article Paul … i agree the flow rate is critical , what is the result if you add better solids filtration , such as a drum filter or plate filter ? can u whiz up some numbers on that comparison?

    cheers

    Reply
    1. Thanks Chris.

      Better solids filtration, such as a drum filter was used in UVI however they removed it because it removed the solids far too quickly so there was very little mineralisation happening in the system. The only change I would make is upsize the solids filters and the bio filters purely from a hydraulic loading point of view. The original design with a flow rate of 167lpm would have done well with the clarifiers, however when they realised the flow rate was too slow they effectively doubled the flow rate and as a result increased the loading on the clarifiers beyond what they were designed for. But it worked for them, but it also increased their washing of the netting in the bio filter section… more work.

      Reply
  2. Chris

    thanks Paul … by increasing the size of the solids filter / bio filter as u mention would that not also increase the workload in cleaning the filters ? from a UVI perspective it worked but if you are to pay labour to clean them day after day it will soon eat into your bottom line ….. was the drum filter they used screening the solids to fine ? an answer must lay on how to reduce the labour in this crucial area of filter cleaning ….

    Cheers

    Reply
    1. Increasing the size is only relative to the type of solids filter, in this case a clarifier that requires at least 20 minute retention time. The cleaning time would not change in terms of labour. You could do away with the two clarifiers and just use one or perhaps change them to a single radial flow filter. Then only one tap to drain. The netting filter is where is all falls over in terms of labour.

      I would go the drum filter and mineralise outside of the stream which will get rid of the clarifiers all together and automate the solids removal. A small one with a 60 micron would probably suit the job.

      Reply
  3. Chris

    mineralise in some sort of holding tank? then pump through the veg, but not return to the fish, send back to the holding tank etc , interesting subject this as i find the solids removal and treatment a huge component of the total labour ..automating makes sense to me ..

    cheers

    Reply
    1. Generally automation will increase the capital cost of the start up. If labor is not “expensive” then it may be better to have a more manual system. Something the potential business owner needs to work on in their plan.

      Reply
  4. shaunmavronicolas

    Paul, when you speak about mineralization… what is the primary end result one wants… is it nitrates, or is it all the other nutrients + nitrates? If one is trying to balance nitrates, so rather not have them or have the flexibility to “add as needed” but wants all the other goodness, how would one go about this? I am asking this since we use a small drum filter (as you know) but seldom put anything back into the system as we have sufficient nitrates… are we loosing out on other important things? What might these nutrients be?

    Thanks.

    Reply
    1. That is a tough one shaun. The primary other nutrients will be phosphorous species amoung some micro nutrients. Depending on what is the make up of your feed source the sludge will be made up of quite a few nutrients not in a plant available form. You can denitrify outside of the production loop which will free up some of those nutrients for use, but dependin on how you denitrify, in some cases some of those micro nutrients are used in the mineralisation process. Controlled mineralisation is quite a complicated process and not something I can explain in a few short sentences. Perhaps I will write up some detail on it when time permits.

      Regards
      Paul

      Reply
  5. shaunmavronicolas

    Sure thing, thanks Paul. I had a feeling there was not going to be “simple” answer. It will be a very interesting topic to know more about.

    Reply
  6. Chris

    i second that its an interesting area to expand on ,look forward to it …

    Reply

Leave a Comment

Your email address will not be published. Required fields are marked *