Automatic Water Changing System (build ideas)

[Aftaburn]

I'm likely going to be moving house shortly & want to set up an automatic water changing system for a variety of size tanks.

The new landlord will let me drill through the floor for drainage.

The tanks are at different heights with capacities ranging from 60 liter to 400 liter.

Anyone got any suggestions?

I had thought possibly a drip system for input & could look at sumps along side each tank running an syphon overflow which overflows to the drain.

The idea is to avoid drilling holes in the tanks & for speed & efficency of implimentation.

If I limit the throughput of the system ideally it should change 20% of the total water volume weekly.

I would rather have a slow enough trickle system that chlorine is not a problem although there may be chloramine in the local water supply.

I don't intend to preheat water either.

The tanks will likely end up spread around a lounge so plumbing lengths will probably vary considerably.

[Ira]

If you're only pouring water in slowly and overflowing the excess down a drain 20% weekly is only going to equate to something like a 10% water change or less if you drain it down and refill. You'll have to be pouring in a lot more than that to make it worth the time and trouble.

[Warren]

Hi Ira,

Sorry to contradict you but a continuous water change changing 20% of the water over a week is more like a 22-23% effective water change per week. As the water is changed continuously, the amount changed has more effect.

The maths behind it are a little difficult to represent graphically due to limitations of the posting module on the site, but it's a fairly basic integral term based off the % change and the time.

[Pies]

Sorry to contradict you but a continuous water change changing 20% of the water over a week is more like a 22-23% effective water change per week. As the water is changed continuously, the amount changed has more effect.

The maths behind it are a little difficult to represent graphically due to limitations of the posting module on the site, but it's a fairly basic integral term based off the % change and the time.

I'd like to see that because based on what IRA said I would have to say hes right. Doing a 20% water change over a period of time would equate to some of the water being changed it, being removed.

I hate to contridict you contridicting IRA but your incorrect and IRA is right (in principle).

Pie

[Aftaburn]

The tanks will need to be sterilized in the move & will be set up with fresh gravel so will be prone to new tank syndrome.

Probably I want to change 30% daily automatically with treatments & possibly a heater on full time in a sump with a ballcock arrangement with failsafe drain.

I've got 20L or 40L plastic header tanks, hoses & tubing some basic taps.

The system has to be able to run a total working aquarium capacity of 1000L.

There are 3 water conditioning agent controls needed.

Has anyone done this or have any ideas on what taps & controls I can use?

[Ira]

I wouldn't have a ballcock or float arrangement or anything like that in the sump, just another overflow there running to a drain should do and you control the water change rate with a valve running fresh water into the tank. IMO, it's more foolproof that way. Unless I'm misunderstanding what you're talking about?

[Warren]

I'd like to see that because based on what IRA said I would have to say hes right. Doing a 20% water change over a period of time would equate to some of the water being changed it, being removed.

I hate to contridict you contridicting IRA but your incorrect and IRA is right (in principle).

Pie

So you're not willing to trust someone who got 98% school C maths, acredited UE and a maths degree then... Believe me, if I tell you it's higher it is... Just for you though, I'll do the workings in Matlab and do a screen dump as a graphic I can put here...

[Pies]

So you're not willing to trust someone who got 98% school C maths, acredited UE and a maths degree then... Believe me, if I tell you it's higher it is... Just for you though, I'll do the workings in Matlab and do a screen dump as a graphic I can put here...

Hahahaha your joking right? The president of the USA told me there were Weapons of Mass Distruction in Iraq and he is the "leader" of the "free world", people get things wrong.

I don't have a Math degree but I do have Science degree. And my science degree and C+ in school cert maths says your wrong.

Pie

[chimera]

I think the important factor here is when the "continuous" water change is performed. You are both right but arguing over different "time periods".

If water is "continuously" changed during the week (eg: 24x7) then Warren is right, the "effective" water change OVER ONE WEEK would be a little over 20%

However, if 20% of the water was changed continuously ONCE per week (ie: water goes in at same time water comes out) the effective water change would equate to LESS than an a 20% change because of some water going in going back out again.

From how I've read the above, it is a misinterpretation or misreading of how the water would actually be changed by the two of you!

[Warren]

Ok, so we don’t get any misinterpretation, here's what my argument is based on;

You normally setup a continuous system to be exactly that, continuous.

A 20% weekly water change is normally done all at once by removing water and adding more (effectively diluting to approx 80%).

This is the method my argument is based on and it's still valid. I'll back it up with the calcs in a day or so when I get time to do it graphically. Incidentally, I passed my calcs past 2 of our PHD engineers and one ME and all agreed... We also modelled it in Matlab with the same result.

The only point left not clarified is where the water enters and exits. All new water enters into the system so it travels through the tanks first then overflows down the drain…

[Midas]

The only point left not clarified is where the water enters and exits. All new water enters into the system so it travels through the tanks first then overflows down the drain…

so you're assuming plug flow rather than well mixed?

[chimera]

This is the method my argument is based on and it's still valid.

Cool, so if you sell something to me for $10 and I give you $20 in $1 coins continuously over the next week, you will give me $15 in change?

[Midas]

Assuming that the system is well mixed, I calc a dilution figure of 18.13% if 20% of the water is changed continuously rather than all at once (which would result in a 20% dilution).

In reality the tank system is probably neither going to be well mixed or plug flow. Are you including any dispersion terms in your model?

Might be easier to measure experimentally than to calculate. Set up a tank with some sort of tracer chemical in it and measure the dilution over time as the water is changed continuously. The experiment could be repeated with different mixing regimes/flow patterns in the tank to see the effects.

[Pies]

I can see this will go nowhere, into and argeument in symantics over 'benifit'. Without presenting the long division (i'll leave that to the math experts) think on this:

You have a 1000 litre tank. It has 100ppm nitrates. If you do a 20% water change, the nitrates will read lower, than if you did a 20% water change slowly over a 7 day period. Assuming that you are adding no nitrates of course. The water you change in (the good stuff) will be removed, thus weakening its effect.

Same thig if you take a 5th of burbon, and do the same experiment, the 20% change will leave less burbon in the bottle than if you added 1% in, then shook the bottle, and took 1% out - repeat until 20% is added.

At least thats what my C+ tells me.

Pies

[Alan]

And what about evaporation?? :roll:

Alan 104

[Warren]

Cool, so if you sell something to me for $10 and I give you $20 in $1 coins continuously over the next week, you will give me $15 in change?

No, but if you give me $22 in $1 coins over the next week I'll give you $10 in change... (Interest mate!!)

[Aftaburn]

Thanks guys... lol... ok here is how I see it...

For starters I am a slug at maths but... 1+1=11

Chimera... I think your maths works the best... makes perfect sense $15 change sounds about rite (i failed englisssh two).

As for the mixing etc... for arguement sake lets say its one 500 liter tank... no my friend who can tie both his shoe laces unassisted says its impossible to put more water in than you put in. I think he might have a point (really not sure its a tough one).

Ok something I need to clarify... the tanks are NOT connected in series for those unsure.

Water input... probably near the filter intake.

Water overflow... from near the surface at the opposite end of the tank.

Mixing very well.

I take it no-one here has actually done this before??

System Definition: Semi-Open

i.e. I think its self explanatory but... partly open system means fresh throughput plus recycling processing simultaneously.

Evapouration issues: None I have excellent lids (hatchet fish) therefore evapouration = condensation and returns to tank (in other words not an issue).

Maths aside any ideas?

To be perfectly honest the mathematical side is a minor detail imho.

The bottom line still has to be if enough is being changed it is if not its not, so... make damn sure it is enough which should be relatively easy.

These tanks have rainbows so effectively changing a huge percentage weekly will likely be appreciated, I'll change heavier rather than lighter the limiting factors being heating and water treatment.

So far as overflows go on a mixing sump... system output is the factor that drives input hence the ballcock arrangement... good point though I could use overflows & just control the overall system input but... that doesn't give precise control over the individual tank inputs the way I'm looking at controlling it.

[suphew]

I'm not sure why this is has become so difficult, it seem pretty simple to me and yes plenty of people do it, alot of ponds also, and alot of larger fish shops use similar (animates in Wlg for one).

What you do is have an over flow pipe, this sets the water level in the tank, you then have a feed from your top up supply with a value on it to each tank, this could set the water flow by either (or both) a timer or a restricter (i.e. only open the tap a little bit), put a bucket under the outlet to messure/set the number of liters per hour (or per day) this gives you your base setting. Try it out, if you find you need more open the tap or increase the timer. And look not a calulator in sight!

[Aftaburn]

Thanks suphew

Any idea how they get around water treatment issues such as fertilizer & water conditioners.

That was basically what I was looking at as an initial system.

Ideally eventually I'd like to move to a PC controlled system but it seems likely quite expensive to set up.

PC benefits include data logging giving a conditions picture over a period of time, also condition/quality triggered events eventually.

[suphew]

Yeah I looked into this for a while to, and found to too expensive. The best option I found was the systems they sell at hydroponics shops they have sets of dosing pumps and controllers for them. You could also use a dosing pump and timer like marine guys do for kalk topup, these run to $100+ each.

Of course the simple option is to dose your top up water, but this doesn't give you control for each tank unless you run multiple top up tanks.

[antwan]

Any verdict on this one?

[Aftaburn]

Yeah I did reach a verdict... gave up on auto.

Couple of reasons...

Firstly if you add 20% on the overflow system the effective change is less than 20%.

Then due to differing heights using 4mm input hose & 19mm overflow I found some tanks got more water than others (even though I thought I had gotten round that, part of the issue could've been the overflows are all connected for 3 tiers of tanks).

Some of the overflows struggled with the capacity (input was 19mm split 12 ways).

Output was 19mm drain... too small, if you use 4mm inputs like this go for at least 25mm output drains with anti clog devices (I have for 19mm but they clog fast).

Water Treatment... all the dosing pumps I looked at were unable to dose at a low enough dose rate.

Plumbing... couldn't plumb the drain permanantly or plumb the input water & automate (most of this was the landlord wasn't keen).

Detritis buildup... with a trickle overflow system you still need to vaccum the tank bottoms anyway.

I did hear of auto systems which would handle the job but the base control unit was around $2000 & would've needed further money spent on extra bits.

Final system currently in use.

I have a 240 litre plastic sump upstairs.

19mm feeder hose to downstairs tanks.

15mm (after mixing tap set & water filters) hose inputs to sump with ballcock emergency cutoff valve... I was unable to plumb a 2nd emergency sump overflow (same deal as toilet cistern) due to plumbing constraints although this would be adviseable... as a result I have to watch very closely when sump filling.

Meathod:

Fill sump via tap arrangement.

Treat water in sump manually.

Manually syphon tanks with hose long enough to go outside or to other drainage.

Fill from preheated sump.

Bottom line... takes a while to fill the sump (25 mins) so if changing over 500 litres at a go is time consuming.

Hot water cylinder can only heat a limited number of sumps plus showers.

Water temperature during sump filling needs careful monitoring (no biggie gotta be handy to watch water level anyway).

Syphoning tanks individually is much better than overflow.

19mm feeder hose takes 17 mins to drain sump fully.

Running up to 20 tanks with this at a capacity not exceeding 2500 litres overall you can do it like this at home... if you doubled capacity it would start to pressure this system & at 50 tanks & over 5000 litre capacity I would suggest you really might want a more serious system... one thing I do like the idea of is if you have 5000 litres of tanks having a sump that is minimally 1000 if not 2000 litres would be nice... also at this plumb the input water & put a heater in the sump along with a circulating pump & automate refill.

Still using a manual filling system at this level would be fine although I would look into 25mm refill hose with a diffuser on the end to prevent tank destruction due to excessive flow rate.

Also its worth noting to spend more money on taps / valves rather than buying cheapos cause you'll get leaks eventually with a pressurised system.

I have 3 taps between sump & end of fill hose... 1 at sump (serious valve) 1 at midway down the stairs (disconnection point for cleaning) & 1 on the end of the hose which would be better if it was a squeeze type than the gardina turny irrigation type that it is (leaks very slowly).

With 2 of the 3 valves closed I have a solid safe cutoff but I prefer 3 in case of one being played with or bumped accidentally.

When not in use I drain the sump to reduce pressure on hoses which also works as another failsafe against accidental overfill & valve failure.

Plumbing down the front of tanks is accessable but ugly.

I never lift water by hand.

This system isn't 100% perfect but it is relatively cheaply built (don't skimp on hose quality I use strong foodgrade clear hose, garden hose doesn't cut the mustard).

Do not pre fertilize your sump as any nutrient will encourage problems growing in it & your plumbing (had issues after treating with peat).

For now... I'll go with the current system it still beats using buckets by 10,000%.

If I ever set up more permanantly & larger I will be going for the supersump. supersump.

:roll:

Anyways hope thats vaguely useful.

[Aftaburn]

Still more dribble... yeah I know I could've individually valved the input lines to ensure even distribution of water (even have the valves) but still is overflow & when I'd built it originally I was certain the octopus I came up with was fail proof (still think it was pretty good as all the outlets were 1 foot long going into larger sheaths which directed to the tanks without adding any pressure issues to the octopus).

At the end of the day... the system currently in use works pretty well.

One thing I did find useful was on a rotational water change system plotting which tanks were done when, helps you keep track & keep on schedule.

[maran]

I hae been wanting set up one like that for a wee while.

Aftaburn, you think you can draw a picture on what your set up like. My small brain cannot fgure out fully what your set up is like ( so I can copy it ).

I was more leaning toward a overflow system. Only trouble is finding a valve that regulates inflow reasonably accurately-you know any place where I can by one and what is it called?

[TimTam22]

I'm not sure why this is has become so difficult, it seem pretty simple to me and yes plenty of people do it, alot of ponds also, and alot of larger fish shops use similar (animates in Wlg for one).

Animates had a accident one day and it pored COLD water into some of there tanks and they lost over 220 fish. It happened over night and they came in the next morning to ALOT of floaters. My ghost knife i currently have lived through the incident as the same thing happened to his tank but it was during the day so was caught in time.......apparently the tech had been in and fiddled with something he wasn't meant to. ANYWAY thats just a bit of useless info for everybody :lol: