Backyard Aquaponics
A place to discuss aquaponics

Hi All,
My little indoor experiment is going well so far. The lettuce seed is germinating, and the nitrogen is cycling. So while I watch my little setup and learn, I'm scribbling ideas for a bigger system--maybe not big by some standards here, but my goal would be to get 15-20kg (liveweight) of fish a year. Plus veggies, herbs and maybe flowers, too. I'd like to share my initial ideas and get your input.

I can get Rubbermaid stock tanks here easily, and have some old steel workbenches. My thought is to use three 50gal/190liter stock tanks for grow beds, and put them up on the workbenches in the sunroom. They are 12in/300mm deep, and I would fill them with gravel and/or expanded clay pellets. For the fish tank, I would use either a 100gal/380 liter or 150gal/570 liter Rubbermaid stock tank directly on the floor. It would be 24in/600mm deep. I'm working inside on a floor over crawlspace, and don't want to cave in the floor by going larger than this. Here's the tricky part: I want use a submersible pump to push the water up to an irrigation timer with three solenoid valves. The water would be routed to each of the 3 grow beds in turn for a pre-set time. That way, the water level in the fish tank doesn't fluctuate as much as if I were pumping water to all of them at the same time. As each bed fills, the water triggers an auto-siphon which has a greater capacity than the pipe carrying water in. That way, if the water is still flowing in as the bed empties, I won't get wet floors. (I think I'm going to -need- to have overlap of the in-flow and out-flow period to make sure the auto-siphon triggers.) Then as the first bed is draining, the next bed begins filling.

For fish, I'm thinking Yellow Perch or Channel Catfish, but am open to suggestions. Here's a link to the Hatchery near me http://www.delmarvaaquatics.com/ so you can see what I know I can get. Although Tilapia is very interesting, I don't know how to get them.

Ideas? Suggestions? Anyone doing something similar?

--Janet.

Hi janet make shore you have an overflow in each bed because solenoid valves aint fool proof un less you pay big bucks

Oh, but Janet, think of the fun you could have with all your fish in the crawlspace!




!

Flooding the crawlspace would regrettably result in flooding the basement. Oh....wait....

--Janet

Janet,
like the idea of the sol valves and timer idea - I am trying to do away with the timer and have liquid sensors to register the water level and switch the solenoid valves.

As long as the auto siphon volume capacity is larger than the pump then valve malfunctioning will be a minor irritation and not a major headache.

What size pump are you planning on using (GPH or GPL figure will do)

Hi Ell,
I haven't really figured out pump size yet, and that's probably a good thing to discuss. I need about 1 meter of head. So the question is, how long of a cycle should I aim for? If I want to fill each bed in about an hour, once every 3 hours, then I think a 10w pump like this (140 gph at 0 head) http://www.thatpetplace.com/Products/KW ... mdy00.aspx would do it. I actually have one of these, so could test with it. Do you think I need to go faster/more frequently? or slower/less frequently?

As far as the idea of eliminating the timer goes, it occurs to me that I could put a sensor on the outlet to the auto-siphon or a float-valve in the growbed. I like the idea of sensing the out-flow of the autosiphon rather than the water level. It seemed that it would require less fine-tuning that way. I wasn't quite sure I could pull that off, though. Even though I am the technical sort, valves and switches are not my strong point.

--Janet.

Janet,
With pump sizes the general opinion is that it depends on water volume and fish densities - if stocked densely (6kg -> 100L, think that is 13lb to 23gals?) then water volume turnover should be 3 times per hour but with lesser densities, less oxygenation is required etc....

The next point to consider is the pipe size of the siphon, water flow into the grow bed has a minimum figure, if you have a read *here* there is a guide to minimum/maximum flows for pipe sizes used (in metric)

Seeing you have the pump, my suggestion would be to set it up with one grow bed only filled with medium (no fish required) and give it a run and make sure the auto siphon works with continuous flow to the bed (hint: bucket water from fish tank to grow bed to cut down on cycle whilst experimenting) - time one full cycle for a reference.

Cycle lengths - mine is a lot shorter than that and I don't know what the optimum is but this is the slowest configuration that I know of *Mike's setup*

Liquid Sensor - this is very basic configuration, bare wires go into the water , circuit is made and globe shines, change battery for power supply required to energise solenoid, swap light for solenoid windings and have the 2 bare wires sealed in the outflow pipe - something a little more sophisticated (but cheap) is *here*, there should be something similar in USA.

That should give you some light reading, don't hesitate to ask for clarification

After thought - the irrigation sol valves sold in Aus work with a minimum input pressure of 70kpa (equivalent to 7 metres of water pressing on plunger), I'm not sure what that equates to in GPH/LPH so buy and try one 1st or check what is avail with lower pressure requirements

Hi Ell,
In the (archaic--grr!) English system, I was working with 0.5 pounds of fish to 1 gallon of fish tank to 2 gallons of filtration as being optimum high-density. ( My proposed system has 1 gal fish tank to 1.5 gal filtration, but I can only fit 3 growbeds in the sunroom.) Outside of optimum, I was assuming that the filter volume would limit my fishload, but it makes sense that turnover is a factor here, too. If I assume 100gal/380 liters in the system, then I have to be able to move water at 300gph/1200lph and 1 meter head. Did I get that right?

Given the limitation of solenoid valves, does it make better sense to run this without them? I've been studying your system write-up, and think that might be what you are doing. Also, I think Simmo is doing that, just with many more beds and more tanks. Ie, run continuous flow to all beds simultaneously. Start the system with different beds filled to different levels so that they discharge via auto-siphon at different times, thus avoiding large tank fluctuations. The vulnerability is that due to natural variation, your beds might sometimes synchronize, meaning greater water level fluctuations in the tank, but then they would drift back out of synchronization. The strength is that it avoids potential technical failures with solenoids, timers, and water sensors. KISS -- Keep It Simple, Stupid.

--Janet.

Well, wait a sec...this solenoid valve seems to handle lower pressure:
http://www.mrdrip.com/sprinklervalves.htm
Is this what we are looking for?

Les, you cracked me up with the patent pending.....................

I don't think the water would flow enough current to actually light the bulb, and in any case you wouln't want it to, as you get electrolysis happening.........degredation of the electrodes unles you use carbon, and conversion of the water to O2 and H2

If any one goes the way of sensors, let me kow, as i can draw up simple circuits for you. Les, in that style you'd be much better putting in a single transistor into the circuit, 50 cents, and it will work with only a millia amp or so of current, while having an output sutible for a digital input of a plc or smart relay.

Float sensors are also easily made, an alarm reed switch from dicksmith, a hollow tube, two cable ties, one tiny magnet and a float made of styrene.

PS smart relay bloke due here in about 40 minutes, so will corner him.

Steve

Bet you will, just give him a drink!

LOL, nah, i'd need to cable tie him to a chair to get him to have a drink while he's working................... LOL

JP,
How were you going to trigger the solenoids on and off? presumably via some sort of timer. I had trouble trying to find a timer with more than 14 cycles in a day. I was working on the flawed assumption that I need to have 3 cycles per hour. Joel advised that his large system only goes on once every 3 hours. Thus I have now restricted my cycles to just every 2 hours.

I also have multple 100L grow beds and was trying the Simmo style autosyphon, but found that it drained too efficiently (too quickly), so changed to having a stand pipe with 2 10mm holes at the base of the pipe which is cut to the height that I want to fill the grow bed to. Any overflow goes down the centre of the standpipe. see second picture on this page http://backyardaquaponics.com/forum/vie ... c&start=15

Make sure with water sensors that the two points are made from stainless steel, they only need to be 1-2 cm long bits of 2-3 mm rod. Have them ~5 mm appart and makesure they are not in contact with gravel (some gravel conducts). The water arround them needs to drain quickly so it registers properly. Ideally use the light activate switch kit from DSE or Altronics/Jaycar and instead of the light dependant resistor put in the water sensor. You may have to have a secondary relay to handle the current of the solenoid, ie 30A horn relay. I used this method for creating bilge pumps in model boats, they are reliable and inexpensive.

Janet,
"Operating pressure: 10 to 150 psi (.7 to 10.5 bar)" 0.7 = 70 kPa which I now know = 10psi

Steve - I agree, it was to give people a picture of what is happenning, but really don't need anything to sophisticated to sense water/no water

Tim - hoping someone would jump onboard with a simple cct, one of the ideas is to place the leads in the outflow pipe to detect the auto siphon kicking in hence no gravel problems