Build an aquaponic indoor farm – part 3 – build log

This is a description of how we at Johannas Stadsodlingar (urban farms) and Concinnity together have built Johanna’s aquaponic pilot facility. We want to share how we did it and our thinking behind it. There is quite a lot to think about, so there will be several posts to cover most things.

There are quite a few parts to the work of building a pilot plant for aquaponics. When we started, we thought it might take 9 months to do the job. We were wrong. It took almost two years. We worked mostly in the evenings and weekends during the first 9 months. After that, one of us started working full time in the company and during the past year we have also had additional help full time. We did not spend all the time building, but a large part of the time was spent on construction.

The building we are in a is an old farmhouse for cows, built in the eighties. It is 900 squremeters, and about 55 meters long and 16 meter wide internally.

IMG_7869

First we had to remove old equipment, such as the milk tank and the traverse for the feed wagon. Below, it is Micke who cuts the I-beam for the traverse where we would build the wall.

IMG_8650
IMG_7881

We chose not to use the entire building for the pilot plant, so needed to build a dividing wall. Here we see when Thomas, Micke and William are working on it.

IMG_8668 12 mikael thomas william wall building

The building had not been used (other than temporarily as a warehouse) for almost 15 years, see we needed to clean everything: walls, roof, cable ladders. It was a lot of work. Here Anke Johanna cleans the ceiling.

IMG_8684

We needed to inspect the drains to make sure they did not have any serious leaks. The drains are used to carry faeces from the fish to an external manure tank. Here Rasmus helps us to inspect the pipes.

IMG_9374

In the drain we found a toad that seem to live there during the winter.

IMG_E9376

We had decided to have an external sump tank. We needed to know that it could be dug at the location we chose and that it would not reach the bedrock, so we enlisted the help of a neighbor who dug a test hole with an excavator. The ”test hole” became large enough that we could improve it a bit by hand (William digs and Micke measures where to put the building) to later use as a hole for the sump tank. The hole was not really straight, as you can clearly see later.

IMG_9051

Work went a little slow so before we could work more on the sump tank and the building on top of it, the winter arrived and the hole was filled with groundwater and froze.

IMG_9557 We built a sump tank structure in wood, which Mikael was happy with, that we later dressed in a rubber lining. IMG_9339

The room that previously had the stall for the calves needed to be furnished to have a table for sowing and the small plants. We call it the cutting room or the calf room. First we needed a floor so we could install new drains.

image_from_ios

Here Mikael builds a table for the small plants, that will stand in water in trays.

IMG_9828

We thought for quite some time about how to hang the LED lights. In the end, it was a fairly simple construction in wood that actually works really well. Mikael came up with a good design in the end.

IMG_9866 IMG_9954 IMG_9960

The tables being covered by rubber lining.

IMG_0034

Building the large troughs for the plants was quite a lot of work, as the floor tilts quite a lot. We have a 10 cm difference between the highest and lowest point on the troughs at a distance of five to six meters. Each piece must be sawn perfectly and measured in place with a laser.

IMG_9898

We bought rainwater tanks with custom fitted rubber lining as fish tanks. Two tanks that hold approximately 9.5 cubic meters. We do not recommend that choice, but more on that later. Here Mikael cuts a hole for the outlet.

IMG_9924 IMG_9928

Piping with 110 mm PVC-U pipes from the fish tanks to the cultivation troughs. There you can also see the evacuation well that we created at the lowest point in the event of a flood. It turned out to be good to have in the future on several occasions.

IMG_0110

In parallel with this work, work was also underway to build our sensor system. Here, Stellan shows the first sensor hub, which measures slightly different things in the cutting room: light, temperature, humidity, carbon dioxide. We later added a sensor to the electricity meter. It also acts as a hub in the alarm system and has, among other things, an SMS modem.

IMG_0131

Here Gabriel is inspecting the first data packets that arrive to our data visualisation system which is based on Grafana and Graphite.

IMG_0136

Now we needed to connect the water system inside the building, fish tanks and cultivation troughs to the outside where the sump tank will be located. Mikael drills holes in the wall.

IMG_0237

Here you can see that our building is well insulated. We have about 15 cm of cellular plastic between the inner and outer wall.

IMG_0240

Then we started constructing the building where the sump tank is located. We needed to cast pillars on which the foundation will rest. It was really shitty weather when Mikael and William made the holes for them with a drill.

03 wet wet wet sump work

We prepare a stable and insulated base for the sump tank.

IMG_0257 IMG_0267

Then we lay the rubber lining that should be on the outside of the sump tank. (It’s made of wood, so it needs to be protected.) We got good help here from Stellan and Gabriel.

IMG_0271

The sump tank is put in place and then we put on the internal rubber lining.

IMG_0272 IMG_0277

The sump tank gets insulation. We also cover it up so that it doesn’t fill with rain until the roof gets built.

IMG_0296

Filling with gravel and soil around the sump tank. William drives his large back loader from Volvo, which we are very happy to have had access to.

IMG_0302 IMG_0358

A timelapse of pouring the concrete for the foundation of the sump house. The whole gang came out that day and helped.

Timelaps of pouring concrete for the sumptank building

Then we started to build the platform, or pier as we call it, around the fish tanks, on which the filter tank will also stand. We see William adding floor boards.

IMG_0332 IMG_0347 IMG_0400

The concrete floor that we have in the building is more or less in direct contact with the ground below, so we wanted to insulate the cultivation troughs. In addition, the rubber cloth becomes less sensitive if you have to walk on it if there is not a hard surface underneath. Here we have a timelapse of installing insulation, the rubber mat and the air hoses in one of the troughs.

Johannas Husby Farm Aquaponics, liner installation

We have installed air stones in the troughs to make sure that we do not get stagnant water and enough oxygen for both the bacterial culture and the plants’ roots. It is possible that we went a bit overboard with the amount of stones.

IMG_1277

We also had to replace all the doors, which we got second hand.

IMG_0775

We insulated the fish tanks at the bottom, so that they came at the right height for the outlets and pipes. After the rubber lining was installed in the tanks, we covered them to avoid getting building dust in them.

IMG_0766 IMG_0891

To reduce the humidity in the main growing room, we chose to install a dehumidifier. It will be more expensive in the long run than installing proper ventilation, but for now we can make do with a dehumidifier.

IMG_1457

The it was time to build the sump tank building. That turned into many a late night.

IMG_0825 IMG_0962 IMG_0977

Installation of the air pumps. It was good that we had a well-insulated sump tank building to install them in. They make a lot of noise.

IMG_1141 IMG_1207

The electricity network where our facility is located is not good. So a backup generator is necessary. But, since we work with live animals, we must have reserve power anyway. Here comes an insulated container for the backup generator.

IMG_1310 IMG_1565

To maintain the right temperature in the water, we have an air-to-water heat pump. For emergencies, there is also an electric heating cartridge installed, which the reserve generator is able to handle the load from.

IMG_1496 IMG_1500

When the weather got better, we were able to finish the roof of the sump tank building (Tomas working on it) and paint it (Anke).

IMG_1731 IMG_1586

We also installed a plexiglass lid on the sump tank so it is easy to see that the air pumps work and that the water level is where we expect it to be.

IMG_1665

Detail of water piping, air hoses, water meter, electric wiring etc.

IMG_1620

When we started filling the system with water, it turned out that we did not do the right thing when we glued the pipes. Many of the pipes needed additional gluing before we got rid of all the leaks. It took several weeks of work. Tip: When gluing PVC-U pipes, have plenty of glue on both pipes. Rubber lining bushings must have suitable glue properly applied.

IMG_1701

Our first attempt to feed the pump for the drum filter used a 120 litre barrel and automatic filling with well water. We had problems with the drum filter flushing at the wrong time or for too long. For that the barrel did not hold enough water. We redesigned this completely. At first we switched to taking water from an IBC tank, but in the end we did what we were supposed to do from the beginning: take water from the clean side of the drum filter.

IMG_1657

We had to install a max water level sensor on the drum filter to regulate the flushing pump.

IMG_1743

When we were satisfied that the leaks were fixed and that the filter flushing worked well, Lisa added bacterial media to the bacteria tank (Moving Bed Bio Reactor – MBBR).

IMG_1837 IMG_1812

We fed the bacterial culture with a little ground up fish food, but mainly with food grade ammonium chloride in powder form. The ammonia was added to a bucket of water the day before it was to be used and then bubbled with an aquarium pump for 24 hours to get rid of some of the chlorine. When the nutrient levels started to rise (ammonium converted to nitrite and nitrite to nitrate) we started to grow plants in the troughs.

IMG_1898

Even without fish in the system and very low nutrient levels, we saw surprisingly good growth, which was even better when the fish arrived.

IMG_1937

But eventually the fish came to our system. Here Tomas talks to Peter who delivers our smolt. We use rainbow in our system from 50 grams to 1.2 kg.

IMG_1928

The sensor system measures a number of different parameters every minute. Here you can see some.

Screenshot 2021-01-10 at 15.36.20

The sensor system delivers data locally to a small computer, but also to our cloud service, so that we can access the data on our mobile phones around the clock, without opening the internal system to the outside.

IMG_2469

The sensor box in the sump tank house also talks to a weather station that sits on the roof. (Even when it is zero degrees outside, we have 30 C in the sump tank house, it may have been a little too well insulated …, the heat comes from the air pumps.)

IMG_2481

Inductive water sensor on the outlet pipes to the fish tanks. It is one of the systems that can send alarms directly to our mobile phones if the water stops flowing in the pipes.

IMG_2468

Automatic water filling for the sump tank. We take system water to rinse the drum filter. That water goes to the manure tank. The fish feces and that water eventually becomes manure for the fields around the farm. Now water recovery in our system is about 99.5%, which we think we can improve to 99.9% or better when we start processing the fish feces for reuse in the system (remineralising it) instead of making manure for the fields.

IMG_2541

A picture of the system in production. We have faster and better growth of our plants than we expected!

Growth at Johannas Stadsodlingar

In the next post, we will discuss a our sensor and data system.

The text in this posted is licensed under Creative Commons BY-NC-SA International.

Build an aquaponic indoor farm – part 2 – design

This is a description of how we at Johannas Stadsodlingar (urban farms) and Concinnity together have built Johanna’s aquaponic pilot facility. We want to share how we did it and our thinking behind it. There is quite a lot to think about, so there will be several posts to cover most things. 

Part 1 – to start, prior knowledge
Part 2 – design (this post)
Part 3 – building log 
Part 4 – water quality and nutrients testing
Part 5 – Production management

Our pilot facility consists of 290 m2 in an insulated building that was previously a cow barn for about 140 cows. The farm was built in the mid 80’s and was then very modern. We have taken a third of the space available in the farm building. The pilot plant uses three rooms: the large cultivation room, of approximately 220 m2, the seeding/sappling room 24 m2 and the ”packaging room” 20m2. The rest are side surfaces. 

In the pilot plant, we will grow leafy greens, spices and rainbow trout (yes, nitrification bacteria of course). 

The first design was based around IBC tanks and three cultivation troughs with 60 x 120 cm rafts. Outside the building there is an area reserved for the sump tank. The fish would be in IBC tanks. We thought of using a radial flow separator to remove the coarsest solid particles and a filter box for the finer particles. Just as described in The Aquaponic Farmer (see part 1).

One of the first designs of the Johannas aquaponic pilot facility (version 3).
2D drawing of one of the first designs (version 4).

In the floor plan above, you can see that we have already rethought using two cultivation troughs instead of three. A narrower trough for rafts from Meteor Systems and one with more traditional rafts from Royal Brinkman (120 x 60 cm). In the Meteor rafts, we use a smaller substrate plug with significantly less material consumption. In the Brinkman rafts we use plastic pots. We were not sure what would work best in the Swedish market, so we wanted to try both.

Since we have a concrete floor, we can not have part of the filter system below the throughs easily, so we investigated other options for the filters. We chose to have a small drum filter instead. At the same time, we began to wonder if the fish we had decided on, the rainbow, would not feel much better in round tanks. In addition, we realized that the sump tank would need proper protection, so we started planning a building over the tank. Eventually, that building became our engine room as well. It turned out that this was very good, but it was a lot of work.

To have better access to the room where we ship out vegetables, we turned the design so that the fish tanks are furthest from the packing room, while it also has less people moving around the fish tanks (the doors are not there) which reduces stress for the fish. The round fish tanks are also more volume efficient.

Round fish tanks and the design flipped 180 degrees (version 8).

In the final design (below) you can see that we added a ”pier” around the fish tanks. The tanks were so high that it would be better to be able to get up a little when working there, without having to stand on a footstool. At the same time, we wanted to get the filter tank (Moving Bed Bio Reactor, MBBR) up from the floor, as the water level in our design is determined by the outlet on the filter tank as well as avoid pipes on the floor so you do not have to step over them all the time. Despite a lot of thinking around this, some of it came out less than optimal. We have a higher water level than we think is good and it would be quite a lot of work to rebuild it to lower the level. But, as I said, the goal of a pilot plant is to learn.

Final design of the pilot facility (version 12).

The facility has 3782 growing places in the two troughs, approximately 3200 growing places (depending on which trays you use) in the tables in the sampling room. We can have a total of about 7000 plants in the system at the same time. 

The water tanks are 9.5 m3 each. We have an IBC tank as MMBR (1m3). The sump tank holds approximately 4 m2, but we normally use 2 m2 of the space. The cultivation troughs together hold approximately 23-25 m3 (depending on how high the water level we set). Together, the total system uses approximately 35 m3 of water. We also have an IBC tank that will be used as a quarantine for incoming fish and preparation tank before slaughter. 

On the far side of the building you can see the sump tank building (beyond the round fish tanks) and the blue cube is a three meter container that contains a diesel powered generator.

In the next blog post we will document our building process in a build logg.

The text in these posts are licensed under Creative Commons BY-NC-SA International.

Build an aquaponic indoor farm – part 1 – start and prior knowledge

This is a description of how we at Johannas Stadsodlingar (urban farms) and Concinnity together have built Johanna’s aquaponic pilot facility. We want to share how we did it and our thinking behind it. There is quite a lot to think about, so there will be several posts to cover most things.

Part 1 – to start, prior knowledge (this post)
Part 2 – design
Part 3 – building log
Part 4 – water quality and nutrients testing
Part 5 – Production management

Three years ago, we at Johannas decided to start a company that produces food in circular production systems on a large scale. We started with fish and vegetables. The method we are going to use for this is called aquaponics. We have experience from cultivation and to start and run companies up to 100 employees. We have been studying circular cultivation systems for many years. However, we have not run an aquaponics operation before and we know that one of the problems people encounter when they start with aquaponics is that they often build a large-scale facility too early. You have to learn how to run an system which contains a mini-ecosystem with fish, bacterial cultures and vegetables. We decided to build a pilot plant at Husby farm in Vallentuna, north of Stockholm. Husby farm is owned by William, who is a partner in Johannas. We chose to use a third (290 m2) of a former dairy farm, which is about 900 m2.

William and Mikael in the old dairy farm at Husby.

If you are going to build an aquaponic facility, large or small, you should study first. We have three sources of information and inspiration that we can recommend. 

The ”basic course” in aquaponics is obtained from the UN’s agricultural agency FAO..It is a book in PDF format that you can download: Small-scale aquaponic food production, FAO 2014.

Small-scale aquaponic food production, FAO 2014.

Our aquaponics design is partly based on a description from the book The Aquaponics Farmer: A Complete Guide to Building and Operating an Aquaponic System, by Adrian Southern and Whelm King.

The Aquaponic farmer, Southern & King.

If you want to run an aquaponic facility on a smaller scale, say 1000 m2 of cultivation area. Then you might be able to cope with what you learn from the books. But we strongly recommend that you participate in some formal training, and example of this is the one-year course Fish and shellfish farming, which is a distance learning course in Sweden. Fish farming is the most sensitive part in aquaponics, and it can be very problematic if the fish dies. In addition, there are legal requirements in Sweden (and probably other countries) to have appropriate training in commercial animal husbandry.

Picture from the Swedish fish and shellfish long distance learning course.

We have also learned a lot from study visits, including at:

There are many more aspects of running an aquaponics facility that you need to have knowledge of. We decided to buid our pilot plant all by ourselves (with a few exceptions), primarily as we think you learn a lot from it. It is a lesson that is well worth it in our eyes, as we intend to scale up what we do much larger, and then we need to understand why things are designed as they are. In your team, you should have at least basic knowledge of the following:

  • Aquaculture
  • Vegetable growing
  • Water chemistry
  • Ecosystems
  • Sales and marketing
  • Business economics, such as accounting etc.
  • Food safety
  • Technical systems, such as pumps, oxygen meters, cooling, heating, etc.
  • Information Technology (IT) system

In our next blog post we will look at the design we picked and the changes we made along the way in our design.

The text in these posts are licensed under Creative Commons BY-NC-SA International.