PDA

View Full Version : Automatic Water Change Scheme


Swarf
06/04/2008, 09:03 AM
I've read about a lot of different methods for automatic or semi-automatic SW changes here on RC, but I had an idea that seems pretty simple, and I wanted to throw it out there for comment.

Sump with overflow drain set at the water level when the return pump is off.

SW reservoir with small pump on a timer pumps some new SW into the sump. Return pump is running.

At a later time sump return pump automatically shuts down for a couple of minutes to drain the excess.

Sump pump re-starts, and you are back to a normal condition.

Does this sound feasible, or is there some reason that this would be a bad method?

Thanks for looking,
Chuck

coralfragger101
06/04/2008, 10:00 AM
Going along that line of thinking you could make it even simpler.

Have a drain that flows to a house drain just above the normal water line in your sump. Any extra water put into the system would then allow "some" to escape down the drain.

Simply add fresh SW into your main display anyway you see fit. This could be from a reservoir, timer and small pump or even manually.

This would result in "some" fresh SW going down the drain but the percentage is minimal.

stugray
06/04/2008, 10:17 AM
coralfragger101,

That would not work too well. When I turn off my return pump, the sump fills about an extra 3 inches. So... everytime I turned off the pump, water would go down the drain, and when I turn it back ON, the level would drop & the pump would run dry.

Swarf's idea is probably a very good one. Add as much water as you like ( and if you overfill, it just drains out ). WHen you are done adding water, turn off the return, and more will drain out, turn return back on, and water level returns to normal.

Sounds good to me, I might have to try that.

It is also a backup to keep the sump from overflowing if the display siphons to the sump or if the ATO runs amok.

Stu

Swarf
06/04/2008, 10:23 AM
Hi Coralfragger,

If I understand your proposal correctly, you suggest moving the level of the sump drain down to the "normal" level when the return pump is running as opposed to the higher level I propose, when the sump return pump is off. Is that correct?

It seems like that might not be a good idea for my proposed setup. If the sump return pump is ever off (on purpose, or accidentally) you will lose quite a bit of SW; whatever volume makes up the difference between the "normal" high level when the return pump is running, and the high level when the return pump is off, basically everything in the overflows, more or less.

I plan on having an ATO system to make up for evaporation losses, so whatever amount of SW that was lost if the return pump is off would be replaced by FW and dilute the tank salinity.

I should have made it clear that there will also be a source of RODI water feeding the sump automatically, I apologize if that wasn't clear.

Cheers,
Chuck

Swarf
06/04/2008, 10:29 AM
Stugray,
Thanks for the vote of confidence. Off the top of my head I can't see any big problems with my scheme.

BUT, it really wouldn't help too much if your ATO system runs amok, unchecked. The drain would prevent flooding, (good), but your tank salinity would constantly be going down as the ATO ran, (bad), killing pretty much everything.

Cheers,
Chuck

stugray
06/04/2008, 12:16 PM
Swarf,

"BUT, it really wouldn't help too much if your ATO system runs amok"

Actually it might IF the ATO is in the same chamber as the drain AND the chamber is relatively small compared to system volume.

If the ATO sticks on, it will dilute the tank until it is overfilled to the drain level. From there on, as the ATO continues to fill MOST of the fresh water will stay on top of the SW and go out the drain ( unless this chamber is very turbulent ).

Of course given enough time, it will continue to dilute this setup just buys you extra time to catch it than normal.

Stu

Swarf
06/04/2008, 01:41 PM
<a href=showthread.php?s=&postid=12678149#post12678149 target=_blank>Originally posted</a> by stugray
Actually it might IF the ATO is in the same chamber as the drain AND the chamber is relatively small compared to system volume.

If the ATO sticks on, it will dilute the tank until it is overfilled to the drain level. From there on, as the ATO continues to fill MOST of the fresh water will stay on top of the SW and go out the drain ( unless this chamber is very turbulent ).

Stu

Good point Stu.

I'm still trying to think of any more downside scenarios, but I can't think of anything too terrible even if there is a failure.

There is just the regular consequences of shutting down the sump return pump, skimmer might not work right, whatever, but a lot of people shut down their return pump for feeding, so you could schedule the two events together.

I don't forsee negative impact on salinity. The water level in the sump will always be at or above normal level during the operation, so the ATO wouldn't be activated.

Cheers,
Chuck

MMM33732
06/04/2008, 02:13 PM
Sounds pretty good to me. Only limiting factor I can think of is the amount of new water you can add will be limited to the same amount of water your sump goes up when the return pump is powered off. If you try to add any more than that while the pump is still running, it will just spill out the drain. Also wondering how you would be able to time out a pump to provide the perfect amount of new water each time. It may be easier to use a solenoid on a timer instead that would empty a container or new water via gravity so the entire container empties therefore insuring an accurate amount of water changed every time. However that would also require you to mix a new batch every time you want to change water which is half the work of a water change.

coralfragger101
06/04/2008, 02:18 PM
You are all correct. I didn't think things through before posting.

Definately - Don't do it my way.

coralfragger101
06/04/2008, 02:48 PM
I wouldn't worry too much about MMM's restrictions. In reality, you aren't restricted to the amount of water you can add. If you add more then yes, water will drain but that water is primarily "old" water. Some of the auto water changes have water coming in and water going out at the same time because the amount of "new" water that gets pulled out is very minimal.

I wouldn't even worry about shutting the main pump down.

With the drain set at the height of the sump water with return pump off then simply add fresh SW on a schedule. If it isn't enough to overflow - so what? It might be the next time, or the next time, or simply when you happen to shut down the system for feeding, routine maintenance or whatever. This would be really good if that overflow from the sump goes into a house drain. This would also be insurance against a house flood. Better for the water to go down a drain than on to the floor.

coralfragger101
06/04/2008, 02:51 PM
Geez - Am I having a bad day. OK, I take that back. You would want to time shutting down the pump for this method otherwise it would throw your ATO off.

Swarf
06/04/2008, 03:53 PM
Man, it is great to get this discussion going, it is making me think the whole thing through.

MMM is definitely right about the limiting factor of only being able to add a volume equal to how much your sump rises. I guess the effect could be mitigated by adding SW on one end, and draining from the other if you happened to add a bit too much, so the new SW didn't just "short circuit" straight to the drain. Coralfragger is saying pretty much the same thing, mostly old water going out even if you exceed the limit a little.

However you meter the water, either time a pump, or time a solenoid, I don't think it matters too much. I don't have surgical precision in mind, you could dial it in a little by adding or subtracting time.

Geez - Am I having a bad day. OK, I take that back. You would want to time shutting down the pump for this method otherwise it would throw your ATO off.

You had me going for a second coralfragger, but you are right the level has to go to normal for the ATO...

Thanks for all the input.

Chuck

rhoptowit
06/04/2008, 04:09 PM
i dont see how any two pumps can equally pump the same exact amount of water over a given time....unless it were a peristaltic pump.

why not do wat randy does and have a peristaltic pump change 1% of water everyday. according to his study that is an equivalent to a 27% water change a month.

stugray
06/04/2008, 04:35 PM
I think the original idea will work perfectly:

1 - install bulkhead in the sump just slightly above where the water level will be if you turn OFF the return pump & let the sump fill.

2 - To do a water change, just add as much new water as you want, just do it slowly so the old & new have time to mix. Even if the new water fills the sump over the drain line, there should be plenty of mixing so you wont be dumping new water ( not much ).

3 - Once the whole water change is complete, turn off the return pump & let the excess drain out.

4- Turn return pump back on, and it should return to the normal running level & the ATO will start working again.

Gravity feed would be the best for the new water feed, but a small pump work work fine as long as it does it slowly.

Stu

Swarf
06/04/2008, 06:11 PM
<a href=showthread.php?s=&postid=12679817#post12679817 target=_blank>Originally posted</a> by rhoptowit
i dont see how any two pumps can equally pump the same exact amount of water over a given time....unless it were a peristaltic pump.

rhoptowit,
There's only one pump, moving water in, so no need to get two pumps to match up. The water moving out of the sump will be by overflow.

Using a dual channel pump is probably the best solution overall, but they are pretty expensive, so I was trying to think of a cheaper alternative that works reasonably well.

Cheers,
Chuck

Swarf
06/04/2008, 06:23 PM
<a href=showthread.php?s=&postid=12680012#post12680012 target=_blank>Originally posted</a> by stugray
I think the original idea will work perfectly:

(snip)

Gravity feed would be the best for the new water feed, but a small pump work work fine as long as it does it slowly.

Stu,
Yeah, I'm still thinking 'six of one, half dozen of the other' about pump versus gravity feed. Since it is SW, and not FW, there isn't really much danger if a solenoid valve faults, and stays open, and dumps the entire reservior into the sump. (Or the feed pump fails to shut down, same scenario). You just end up losing however much you had invested in the salt mix.

Chuck

Siffy
06/04/2008, 09:27 PM
It's not automatic, but here's my semi-auto approach. The problem presented itself when I was helping bag a lot of corals out of a friend's tank. Overall we removed about 15g from his 120 over a weekend so we were constantly adding NSW to prevent his osmolator from diluting the tank.

My solution to that was to add a simple 3-way switch (which is essentially a DPST switch) after my ATO controller. With an osmolator it would require a relay with the contactor being the same as the output voltage, but I have a Tsunami so it's much more simple. I just put the 3-way switch hooked into the power outlet of the ATO and the 2 outputs of it to 2 different pumps. 1 of the pumps draws in RO/DI or kalk from one reservoir and the other pump draws NSW from a 2nd reservoir.

The other part of the "automation" or simplifying is adding a Tee and a ball valve on my tank's drain. Then all I have to do is flip the top off switch to SW and open that valve to remove water from the system. The same amount I drain out will be replaced by the top off. When it's done I just flip the switch back to normal RO/DI operation. You could either hard plumb the drain or throw a barb after the valve for anytime you need to fill a bucket or bag with water and prevent dripping it everywhere. Kills multiple birds with one DIY project imo.

I've designed it, but I haven't built it yet so no pictures.

Siffy
06/04/2008, 09:34 PM
Oh, for 100% automation of it you could install an electric valve on a timer and a DPST relay rather than the DPST switch. Throw all that on a timer and let it run for however long you want each day.

For instance you could measure how much this valve would drain in X minutes
http://www.fishbowl-innovations.com/product/solenoid

Then the 1 timer would control both "states" if you will. Say at 1am or whenever it puts the solenoid into drain mode and kicks the ATO into SW mode for 15 minutes. Then when the timer turns off everything goes back to normal. The only problem with that is you'll either end up with dilution of salinity (when you use a pump that has a lower output than the valve) or a shortened pump life from it kicking on and off multiple times in that 15 minute span (when you use a pump that has a higher output than the valve can drain). The solution then is to put the pump on a 2nd timer and let it run for a few minutes longer than the valve and dial the pump back to very low output. It would take some adjustments, but it could work.

Swarf
06/05/2008, 08:31 AM
Siffy,
I read through your post a couple of times, and I think I understand what you are getting at. In the most basic sense what you are proposing is a combination of switches, relays, and valves that switch between either a SW source or an RODI source to top off the sump.

My proposal drains by overflow, where yours does not, which requires no devices, and has an advantage: when the new SW is going into the sump, the sump level goes to a higher than normal level, so the float switch for the RODI stays up, and the ATO system never tries to add water. You don't have to de-energize the ATO system to make sure it doesn't lower your salinity.

The crux of it is that normally, when you are adding fresh water you want a net change in water level, and when adding saltwater you do not. If you don't hold to that then salinity will change. There are a lot of different ways to do it by adding more hardware, but adding more devices also increases the number of different ways the system can fail.

The situation you described with your friends tank was abnormal in the sense that you were removing things from the tank, enough corals that 15 gallons of water were needed to get back to normal water level. If you added fresh water the salinity would go down, so you had to sort of break the rules, and change the level while adding SW. (You didn't really break the rule, net volume change was zero, you replaced solid volume with liquid volume).

I'll concede that a twin dosing pump is a great method for adding and removing equal volumes of SW, but at a cost over $200 at least. My method can use one cheap pump, and doesn't rely on any precision, you want to be 'in the ballpark', but you don't have to match pump outputs, or time valves to ensure SW volume in matches what is going out.

I guess the one area where the nethod I propose requires some precision is that you don't want your drain level to be at all above the level your sump goes to with the return pump off. If it is above that level, in the long run, you will be increasing salinity. When your return pump kicks back on your sump would be too high, and would have to lower by evaporation before the ATO kicks in, leaving extra salt.

Swarf
06/05/2008, 09:16 AM
I guess I am replying to myself but...

When I was replying to Siffy I realized where the scheme I propose DOES require precision:

The level of the overflow drain has to match accurately the level the sump goes to when the return pump is off.

If the overflow drain is too high salinity will go up, if the drain is too low the salinity will go down. The amount of salt added or lost per cycle would be whatever is dissolved in the volume of water you are off by.

I don't know if that would be a 'deal breaker', or whether the effect could be dealt with.

Thanks to everybody for all the discussion so far.

Chuck

Siffy
06/05/2008, 10:54 AM
<a href=showthread.php?s=&postid=12684380#post12684380 target=_blank>Originally posted</a> by Swarf
Siffy,
I read through your post a couple of times, and I think I understand what you are getting at. In the most basic sense what you are proposing is a combination of switches, relays, and valves that switch between either a SW source or an RODI source to top off the sump.

Yeah, you've got it. :) Except for my case it's just 1 regular light switch and one more valve on my plumbing. Plus the same pump that I'd be plugging in and unplugging when doing a water change. It just gives me the benefit of letting the ATO decide when to "unplug" and I can walk off for a couple minutes if I need to rather than watch my sump level.

<a href=showthread.php?s=&postid=12684380#post12684380 target=_blank>Originally posted</a> by Swarf
My proposal drains by overflow, where yours does not, which requires no devices, and has an advantage: when the new SW is going into the sump, the sump level goes to a higher than normal level, so the float switch for the RODI stays up, and the ATO system never tries to add water. You don't have to de-energize the ATO system to make sure it doesn't lower your salinity.

I don't de-energize the ATO. I redirect it to a different pump. The switch goes between the ATO and the pumps, not between the ATO and the power outlet. Plus 10 minutes of my top off pump running won't affect salinity in my system much at all. That's still in the milliliters range with my peristaltic pump(s).

<a href=showthread.php?s=&postid=12684380#post12684380 target=_blank>Originally posted</a> by Swarf
The crux of it is that normally, when you are adding fresh water you want a net change in water level, and when adding saltwater you do not. If you don't hold to that then salinity will change. There are a lot of different ways to do it by adding more hardware, but adding more devices also increases the number of different ways the system can fail.

Yeah, a lot of that stuff, if you wanted to go with automated, would cause plenty of anxiety. I fully support KISS, and that's why I'm going with the semi automated version. Household light switches generally take years to fail with daily use and valves rarely leak. Their biggest problem is sticking in position, but if it does leak that's why it's installed over the sump.

<a href=showthread.php?s=&postid=12684380#post12684380 target=_blank>Originally posted</a> by Swarf
The situation you described with your friends tank was abnormal in the sense that you were removing things from the tank, enough corals that 15 gallons of water were needed to get back to normal water level. If you added fresh water the salinity would go down, so you had to sort of break the rules, and change the level while adding SW. (You didn't really break the rule, net volume change was zero, you replaced solid volume with liquid volume).

That was a rare occasion. He's moving at the end of this month so he's getting ready for a tank move. Lots of fragging back to make things easier. And lots of trading in to the LFS. But it wasn't 15 gallons of solid volume, that was counting all the water that went into the bags as well. 15 gallons of solid coral would be insane! :) We just kept having to add 2-3 gallons then bag corals then add another 2-3 gallons. And I just kept thinking there must be an easier way. And there would have been if an Osmolator hooked up to other pumps easily. I just wanted to take it a step further with no plugging in. I hate having to mess with cords and identify them in tangled messes. A labeled switch is so much less headache.

<a href=showthread.php?s=&postid=12684380#post12684380 target=_blank>Originally posted</a> by Swarf
I'll concede that a twin dosing pump is a great method for adding and removing equal volumes of SW, but at a cost over $200 at least. My method can use one cheap pump, and doesn't rely on any precision, you want to be 'in the ballpark', but you don't have to match pump outputs, or time valves to ensure SW volume in matches what is going out.

You can get a cole parmer off ebay for $60-70 and sometimes even less. Stay away from the aquadoser garbage if you (or anyone else) goes that route. I've got an SP3000 single head and I hate it. I've also got a variable speed C-P Masterflex and I love it. It takes up more space and makes some noise, but at least it works without check valves and can move air like a peristaltic pump should.

<a href=showthread.php?s=&postid=12684380#post12684380 target=_blank>Originally posted</a> by Swarf
I guess the one area where the nethod I propose requires some precision is that you don't want your drain level to be at all above the level your sump goes to with the return pump off. If it is above that level, in the long run, you will be increasing salinity. When your return pump kicks back on your sump would be too high, and would have to lower by evaporation before the ATO kicks in, leaving extra salt.

Unless you're proposing to turn the return pump off every time you perform an automated water change then your water level will always be at the overflow level. That is until there's a power outage and the level in the sump increases causing it to overflow. When the power comes back on, your ATO will have to compensate for the loss. But rather than worrying about the placement of the overflow bulkhead you can adjust your ATO level.

I wouldn't want to put that much stress on the return pump on and off all the time. That would lead to premature failure I believe. And as it's the most vital piece of equipment on most systems, it's not one to mess around with.

Swarf
06/05/2008, 06:14 PM
Thanks for all the discussion Siffy, I think we are pretty much on the same page. I'm more fully understanding how your system would work, and I like it.

My proposal you add SW first then the excess drains, and with yours, you drain some SW first, but have the ATO set to add SW instead of RODI.

Either method could be automated, and both methods have failure modes.

For your method if the drain failed to open it would be no big deal, but if it failed in the open position it would be a big deal. Also, you have to have a device (relay probably) to switch between RODI water and SW feeding the sump, if that failed you could run into problems too.

With my method the biggest problem would be if the return pump didn't stop to allow the excess water to drain.

Unless you're proposing to turn the return pump off every time you perform an automated water change then your water level will always be at the overflow level. That is until there's a power outage and the level in the sump increases causing it to overflow. When the power comes back on, your ATO will have to compensate for the loss. But rather than worrying about the placement of the overflow bulkhead you can adjust your ATO level. If you look back at my proposal you will see that I do intend to turn the return pump off every time after adding SW, in fact it is absolutely essential to turn the pump off because I am relying on the change in the sump water level caused by shutting down the return pump to drain excess water. The drain would be at the same level as the water would go if you lost power, so you wouldn't lose any water if the power went out, so the ATO wouldn't have to add any water to compensate.

I wouldn't want to put that much stress on the return pump on and off all the time. That would lead to premature failure I believe. And as it's the most vital piece of equipment on most systems, it's not one to mess around with.I guess that is a pretty valid concern, but I'm not talking about constantly cycling the return pump, but only turning it off for a few minutes every day or two.

Cheers,
Chuck

Siffy
06/05/2008, 11:49 PM
<a href=showthread.php?s=&postid=12688184#post12688184 target=_blank>Originally posted</a> by Swarf
For your method if the drain failed to open it would be no big deal, but if it failed in the open position it would be a big deal. Also, you have to have a device (relay probably) to switch between RODI water and SW feeding the sump, if that failed you could run into problems too.

And I have no idea how reliable or finicky those small solenoid valves I linked to are. Lots of people use them on their RO/DI units and I rarely see comments about them failing. But still, I don't plan to use one or go with complete automation. I will be the biggest point of failure with my design, imo. I could easily forget to set the switch back into RO/DI position. And relays are pretty reliable too. Just don't skimp on them. Most electrical stuff fails from overheating, so the best way to prolong their life is to keep them in a well ventilated area and use one that's overrated for the application. ie, use a 5 amp relay for a 1 amp or less task, and a 30 amp relay for a 5 amp task. I just broke that rule by picking up some really cheap relays to play around with, but they won't be seeing any vital equipment like pumps.
http://surpluscenter.com/item.asp?UID=2008060305541657&item=11-1132&catname=electric
And the base for them is only $2. I'm just going to use them to turn off my lights when it gets a temp too high signal.

<a href=showthread.php?s=&postid=12688184#post12688184 target=_blank>Originally posted</a> by Swarf
If you look back at my proposal you will see that I do intend to turn the return pump off every time after adding SW, in fact it is absolutely essential to turn the pump off because I am relying on the change in the sump water level caused by shutting down the return pump to drain excess water. The drain would be at the same level as the water would go if you lost power, so you wouldn't lose any water if the power went out, so the ATO wouldn't have to add any water to compensate.

I guess that is a pretty valid concern, but I'm not talking about constantly cycling the return pump, but only turning it off for a few minutes every day or two.

Cheers,
Chuck

Depending on the return pump, even cycling it every 2 days may take a toll and cause it to fail within a year. You should contact the pump manufacturer and get their opinion. Companies like Sequence, Iwaki, PanWorld, etc will tell you not to do it. The mfgs will tell you SeaSwirls are [B]never[B] supposed to be turned off, and I've witnessed why. We were constantly tripping a new tank's GFCI due to the MHs and trying to diagnose it. Well, after about the 4th or 5th trip the seaswirl didn't come back on. The tank and swirl were both less than 2 weeks old and turning it off 4 times had stripped its gears.

Swarf
06/06/2008, 08:39 AM
Siffy,
All great information. I was thinking there wouldn't be any problem stopping and starting the return pump, but I guess it depends on the pump design. I wasn't really considering the price of the pump as part of the system, because it would already have been present whether or not I set up some kind of auto SW change or not, but if it contributes to the untimely demise of the pump then that would be prohibitive. I have a Gen-X PCX100 that I'm going to use for a return pump, but I haven't researched the starting and stopping issue.

I have boxes of SPST and SPDT 115VAC actuated relays from eBay, I'm not against (slightly) higher tech solutions to automating aquarium maintenance, but I don't try to add layers of complexity where they aren't needed either.

Really, the prospect of a used peristaltic pump from eBay is looking more and more attractive.

Cheers,
Chuck

macawmagic
06/29/2008, 10:54 PM
this is how i'm setting mine up...hope it helps.

drain line in sump that is at correct running level of water...has a ball valve on it so I only open it when i'm about to do a water change.

waters out in the garage mixing. I ran the plumbing through the attic along with an extension cord. Open the drain valve. plug in the extension cord. water flows into tank. forcing old water down to sump. which then rises the sump level. which then causes the water to drain out. and then once all the water drains out. shut the ball valve. We're talking a 5 min water change for 35ish gallons. and all I do is open/close a ball valve and plug in an extension cord to the pump.

hope that helps!

WLachnit
06/30/2008, 12:12 AM
How do you have the plumbing arranged in the DT for the new SW from the garage?

coralfragger101
06/30/2008, 06:13 AM
Mac's solution is probably the easiest to accomplish however you do lose the "safety net" and automation that other solutions have brought into the picture.

Plumbed one of the other ways if your tank ends up with too much water (for whatever reason) - it will simply be expelled at some point when it hits the drain in the sump. Mac's will not because the ball valve is closed in normal position.

I think I still like Mac's solution the best for an "Easy water change method" but not an "Automatic water change method".

macawmagic
06/30/2008, 07:36 AM
well...when the tub in the garage has rodi I open the ballvalve that goes to my auto topoff bin. and then when I make saltwater in the bin, I open the valve that goes to the tank. I'm sure theres probably a better setup out there...but this works for what I need it to do. aka I don't have to carry buckets of water anymore!