In Sand Bio-filter Construction and Operation -Part 1, we provided the basic guidelines around what is the heart of the iAVs system. In this article, we cover the sand bio-filter in greater detail.
Drainage Performance Variables
The drainage dynamics of any given media are influenced by:
- the media density
- particle size distribution
- pore volume
- particle sorting factors
Water will continue to drain from the bed for as much as 20 minutes after the pump stops. During the remaining drain cycle interval, the sand remains hydrated but not saturated. At the same time, the void spaces between the grains of sand are charged with air (21% Oxygen) – facilitating soil aerobic metabolism which includes plants and hundreds (if not thousands) of micro and macro fauna species.
Container Aspect Ratio
Proportions of width to length (the aspect ratio) are not strictly defined. While any shape (from a plan view perspective) – square, rectangle, circle, oval, triangle, polygon – is possible, the sand container should be sloped toward the drain end to facilitate unrestricted drainage. No water (aside from that adhering to the sand particles) should remain in the bed for the final 90 minutes of the drain cycle.
Enhancing Drainage in Larger Sand Beds
It is recommended that agriculture drainage pipe (with a permeable sleeve/sock to exclude the sand) be used where the sand bed exceeds 2.5 to 3 meters in length. This does depend, in large measure, on the specific drainage characteristics of the sand that you are using.
If the sand bed is 3± or more meters long – but from 1 to 2 meters wide – place the drain pipe along the longitudinal center mid-line directed toward the drainage point. For sand beds wider wider than 4 to 5 meters (regardless of length), the use of multiple drain pipes – spaced approximately 2 to 2.5 meters apart – is recommended. If you construct your bed(s) on ground, one can also partially inset the drainage pipe(s) into half-round channel(s) below grade and/or form slight down slope from the outside bed edges toward the central drain line. In all cases the impermeable liner material is placed under the drainage line(s).
Sand Biofilter Depth
A minimum sand bed depth of 300mm (12 inches) is recommended. While shallow beds (of 200mm or 8 inches) may be suitable for some plant species, this has yet to be tested. In any case, any reduction in sand volume in the bed will result in a proportionate reduction in the micro-fauna in the bed.
Deeper beds may be appropriate for very large plants and/or longer term crops.
Furrows
Furrows are an essential feature of the iAVs design because:
- They allow flooding of the sand media – while allowing the crown of the plant to remain dry.
- They ensure that the hydraulic conductivity – the rate at which the water percolates into and through the sand – to remain intact.
- They provide for rapid distribution of the nutrient rich water along the entire length of the bed.
- They are most stable when the lateral cross sectional profile approximates a parabola. Round over the top edges to avoid having the sand collapse into the furrow trough.
Ridges may be built-up from the sand excavated to form the furrows and/or add additional sand to raise ridges above the original level surface.
Depending on crop spacing, the open upper width can range from 7 to 13 cm (3 to 5 inches), and the total depth from 5 to 10 cm (2 to 4 inches). Place furrows around the entire perimeter and then between each row of plants such that all plant rows have furrows on both sides. Connect/intersect all furrows in the bed to ensure even distribution. You can also place intermediate channels connecting adjacent longitudinal furrows, and between plants within a row, if desired.
The goal is to distribute water – and the accompanying ‘wastes’ – as uniformly as possible across the sand bed surface – without wetting (or allowing detritus to accumulate on) the plant’s crown (stem at ground level). Keep the aerial portion of all plant tissues dry and clean of any detritus to inhibit mold, fungi and other nasties.
Pristine Sand Bed w/ tomato transplants being irrigated
As detritus accumulates – and coats the furrow slopes – they will become increasingly stable (resistant to erosion). Algae growth will also tend to cement sand grains in place. Until some detritus coats the bottom most portion of the furrows, lateral water movement will be relatively inhibited but as you see in the above photo, still does fairly well. Uniformity of lateral distribution increases with time presuming furrow channels remain level, adjust gradient as needed until channels stabilize.
Sand Retention
The drainage velocities are typically not sufficient to transport sand grains to any significant extent. But transport issues you may experience will depend on the orientation and shape of the drainage outlet provided. If using a crack or slit approach, then some medium gravel against (interior to) the slot and covered by a second layer of smaller gravel over the first will retain the sand. In lieu of the large gravel, a strip of 1/4″ mesh hardware cloth (galvanized metal) or a course insect screen (aluminum or plastic) will also work to restraint the pea gravel and subsequently the sand. (BTW, I’ve had 20 meter long sand beds with zero screening or gravel placed at the drainage outlet and did not observe any sand exiting the beds (into the sump, in that instance).
Potential for multiple sand beds per grow-out tank
While the usual iAVs configuration comprised one fish tank and a sand biofilter, there is no reason while a fish tank ought not support several sand beds. Ultimately, the goal is match biofiltration/nitrification capacity to the amount of fish food required for sustainable fish production while providing sufficient nutrients to grow a full range of plants. An initial fish tank to sand biofilter volume ratio of 1:2 is suggested, assuming an adequate stocking density and feed rate, but how this is achieved will depend on the circumstances and preferences of the system operator.
This volumetric ratio may be increased over time if/when fish biomass density and feed rate is increased. The object is to balance plant assimilation/uptake with the ‘waste’ production from a given feed ration. The greater the daily feed ration is, the larger the biofilter (and plants grown) will be needed to balance uptake with input.
If you start at low fish density/feed rate relative to the total available sand filter, then you do not need to employ the entire filter volume available.
You can create a larger filter than initially required and progressively increase the portion in use at any given time – as the fish grow/feed rate increases.
To constrain the filter volume being used in the moment, merely block irrigation channels with some sand and/or insert a temporary barrier into the sand column.
Alternatively, one can add additional individual sand beds when necessary to process additional ‘wastes’. In which case, staggering the tank withdrawal events over time will extract the ‘wastes’ on a more frequent basis.
(click image to enlarge)
NOTE: Illustrations are intended to be descriptive, not prescriptive; representative not definitive. Sand filters of larger dimension will benefit from additional drainage provisions as mentioned.
Irrigation Intervals and Volumes
What worked best for Mark was to move approximately 1/4 of the tank volume (drawn from the bottom of the tank) – 8 times per day at 2 hour intervals. This provides two complete tank volume exchanges daily.
The drainage cycle is such that water begins to drain from the bed well before the end of the flood cycle. This means that fluctuations in the fish tank level are quite modest. Depending on your sand bed pore volume (relative to fish tank volume), the water level in the tank will drop by about 10% … more or less.
The number of events per day will depend on your latitude and the season.
Schedule the first daily irrigation event to begin at dawn and begin the last event of each day so the sand bed drains before dark.
Pump Flow Rate
Sizing your pump’s delivery rate to your ‘system’ size is important.
Factors that influence pump size include:
- the tank volume
- the biofilter pore volume (saturation volume)
- the amount of water that the pump will lift given the distance between the water level in the fish tank and the surface of the water in the sand bed (described as the head). Pumps are rated in terms of the amount of water that they will pump at a given head.
A starting point would be to select a pump rate to deliver a volume per hour at the total effective head approximately equal to the tank volume at capacity.
For example, if the fish tank holds 1000 litres (or about 265 US gallons) – and the distance between the water level in the fish tank and the surface water level in the sand beds is one metre – you will need a pump that delivers 1000 litres per hour at one metre head. You will be able to confirm the correct pump size by referring to the pumping rate chart which is often printed on the packaging that contains the pump.
In the example provided, the likely pump capacity will be around 1500± litres per hour at zero head.
Water delivery onto the sand surface
Typically, the velocity (force) of the water delivered from the pump (via pipe or hose) will be sufficient to produce very substantial scouring (erosion) of the sand where it would contact the surface. It is recommended, therefore, to reduce the velocity of the water at the point of impact.
A simple method is to direct the flow onto a pad, plastic sheet, rock or rocks to reduce the force of the water before it contacts the sand. Perhaps a better method is to use a distribution header or manifold across the end of the sand filter.
Mark found that using a 20 mm (3/4″) hose or pipe from the pump inserted into a 40 to 50 mm (1.5″ to 2″) diameter header pipe, which then had 8 to 10 mm (3/8″) holes drilled at 10 cm (4″ intervals) and facing down (toward the sand surface), eliminated scouring and also evenly distributed the volume across the width of the bed. (see above photo and diagram).
As another example, in a much larger ‘system’, with a 50 mm ( 2 inch) pump discharge, he delivered the flow into a poly-lined trough (trench approx. 30 cm wide by 15 cm deep) running the 30 m length of the greenhouse and from which several dozen longitudinal distribution furrows were coupled (intersected at right angles).
Timers
Use a programable digital timer to control pumping cycles.
The cheap electro-mechanical timers that provide for nominal 15 minute increments are not sufficiently accurate – and nor do they allow for the fine tuning that may be required to provide the optimum pumping duration. Start with a 20 minutes ON/100 minutes OFF cycle time – and observe how the sand bed fills and drains. Adjust the timer cycle – up or down – so that, with each flood cycle, you transfer as much water as possible without immersing the plant crowns.
Cycle Duration
Depending on the actual hydraulic conductivity of your media (none are identical) the ON cycle time might be as little as 10 minutes or as much as 25 minutes.
With well-drained sand, you should start to see drainage returning between 1/3 and 1/2 of the way through the total ON cycle time. Continue pumping as drainage is returning until the furrows are completely full and all the sand is fully saturated. Do not flood the elevated ridges between the furrows. Wetting the base of the plant will produce crown rot.
(click image to enlarge)
-o0o-
Hi Mark/Gary,
Great info on this site, I’ve not gone through it all so apologies if I missed what I’m looking for… and I realise what I am about to ask is a ‘piece of string’ type question, but..
.. are there any guidelines in terms of m2 of sand bed (or m3 if not assuming a standard 300m depth) that are required per g of fish feed input per day. I’m coming at this principally from a mineralisation viewpoint, not a plants consuming nitrogen input perspective. i.e at what point would the sand beds become overloaded with organic matter and possibly become anerobic.
I’d like to convert one my existing systems to IAVS, though would probably do it in stages. I therefore need to have an idea of how much feed (g/day) a standard BYAP growbed (2.2 x 1.2 x 0.3m) converted to IAVS would be able to handle.
I know there are lots of variables eg % protein, type of fish, temperature, though wanted a ball park figure. I know the Wilson Lennard XL calculator has a figure of 45g/m2 of 300mm deep hydroton bed for 50% feed i.e 110g/ BYAP bed though from experience the BYAP beds do not cope well with sustained feed at this rate.
I would assume given the IAVS enhanced mineralisation capability that this figure would be a safe starting point, though how much higher could IAVS handle?
Thanks
Matt
Thanks for putting these all together.. I focus on Aquaponics from India and have been reading McMurtry’s work for several months or maybe more than a year.. After seeing this website, I finally started to feel good about my own humble beginnings with sand wicking beds and gravel media beds with red wriggler worms.. with a very low-watt pump that turn around the system in 8-9 hours..etc. With time my systems got simply evolved by adopting several UVI processes as well. I will need to re-modify some processes.. and you are an inspiration, Gary. I will be following your updates closely.
Have you tried red wriggler worms in sand beds?
VK Narayanan…..The credit for this site belongs to Mark…..he’s the inventor of iAVs.
Composting worms (red wrigglers) will show up anywhere that conditions are right for them. I’ve known them to live in trickling bio-filters and gravel beds so I’d be surprised if they didn’t also inhabit sand bio-filters.
Good luck with your system……and feel free to update us with your progress.
I am considering building at least two sand beds, that will be approx 48″w x 16″h x 10’l. My plan is to have the beds aligned end-to-end, with individual drains and the drains dumping into a common sump to be pumped back to the fish tank. The water level in the fish tank will be approximately 42″ higher than the top of the sand bed. So, gravity flow to the sand beds, pumped back into the fish tank.
I have a few more construction questions:
1. With the size listed above, do you recommend that the entire bed be pitched towards the drain, or will the ADS perf pipe with sock be sufficient to drain (I realize that an inch or so of water may always stay in the sand/ADS pipe) to drain the sand bed properly?
2. Regarding the liner, do you recommend that the long side of the tank have the liner tucked tight to the bottom corners, or is a radius along the bottom acceptable?
3. With the bed dimensions listed, using both 1/2″ plywood walls and an upper and lower 2×4 “band” for reinforcement, does your experience tell you that some ~48″ cross bracing will be required? In other words do you expect that the long walls may try to bow out with the weight of the sand/water?
4. I there a way to send a drawing or post a photo to the comments?
Thanks !!!! mh
RE: 1. I recommend that the bottom of the beds and any drainage slope down toward the return. You do not want a permanently water-logged zone at the bottom to potentially go anaerobic.
2. If I understand the question, the edges where the sides meet the bottom plane can have a radius – as long as not too large of one. OTOH this will make folding the liner neatly at the corners more difficult.
3. Depends on whether or not you are staking the corners/sides into the ground. If not then a cross brace may be beneficial. My plywood beds were side-by-side so the lateral forces canceled each other out. OTOH, you could add a 2×4 or 2×6 top cap (wide face horizontal) which would also double as a sill to step, kneel or sit on.
4. No idea- Gary probably knows
Mark,
Ah….the voice of experience, I appreciate the good advice. Time to dust off the table saw and start ripping plywood.
Gary…..Thanks again for putting this site together.
………mh
I am considering using a cattle feed bunk as a sand bed. I went by my local Tractor Supply store and the feeders appear to be pretty rugged. http://t.tractorsupp…-feeder-10-ft-l Another nice feature is that the liners are replaceable.
I checked with the feeder mfg to try to find out if they could tell me about the type of polyethylene used in the liner, they didn’t offer any details.
http://tarterusa.com…ty-bunk-feeder/
A couple of drain holes could be easily plugged and a uniseal or a bulkhead and off you go. The side walls are about 12″, so I was thinking of making the sand about 10-11″ deep to leave a bit of elevation for the water/furrows.
Thoughts, any concerns……?
1. I seriously doubt those legs will support 18 cu ft of sand (~2000 lbs dry), even if your ground (under those 2 tubes) does.
2. That’s a bit on the shallow side especially with allowance for freeboard.
3. That would ‘give’ approx. 18 cu ft of sand – for only about 60 to 70 gallons of tank volume.
Assuming you have some basic carpentry tools and skills, a box made of 1/2″ CDX plywood with an exterior perimeter frame of pressure treated 2×4 and lined with HDPE, EPDM or Duraskrim would be far cheaper and allow for about any size you would wish (backyard that is). Here in Montana, I could get all the materials for a 4′ x 12′ box at 16″ deep for about $60 to $70 and a food-grade liner for $60 to $80 depending on material and source chosen (maybe another $10 for screws, don’t use nails). That would accommodate a bit less than 2 cubic yards of sand, for a tank of approx. 175 gal. … or about 3 times the size (area and volume) for about 1/2 the cost.