Planning: auto water change system

[jeremyjoslin]

Looking for some feedback on my design before I build it out.

Here's the walk through:

1.) RO/DI water enters 50 gallon Brute #1 and begins to fill. Flow is regulated by a solenoid wired to Apex and is closed in fail state.

2.) As water fills Brute #1 it overflows into Brute #2 and fills that.

3.) Eventually top float switch gets triggered and tells Solenoid to close turning water off (now both are filled with fresh water).

4.) I get an email alert from Apex telling me that the bins are full and I can walk over and dump in exactly one bag of salt to get desired specific gravity. Note, I'll make sure I place top float switch to try to match the exact 44 gallons of water needed to mix FSW at 1.025sg.

5.) The powerhead at bottom of Brute #2 will mix for the one-two weeks between water changes (timing TBD). A heater will keep the water at system temp. A dedicated temp probe on my PM1 will regulate this via Apex.

6.) When the time to change out water comes (programmed automatically via Apex for every 2 weeks, for example), a submersed pump in my sump will turn on and dump water into a drain until a dedicated float switch in my sump triggers "water change low level" and stops it. This quantity should match the 44 gallons in Brute #2.

7.) When the above float switch measures the low level, it will trigger the Mag 9 to start pumping water from the FSW Brute (#2) into the sump.

8.) The pump will continue until the lower float switch in Brute #2 gets triggered which signifies a low water level in Brute #2, OR until my master float switch in my sump gets triggered indicating a "full sump" (same one I use for ATO and my "master" water level). Note: I'd also like to include a failsafe with my Apex such that it knows I've added salt to Brute #2 and won't dump 44 gallons of fresh water into the sump if I've forgotten to add salt for some reason. Can this be a manual step or acknowledgement somewhere???

9.) Once Brute #2 is done dumping its payload, the Mag will shut off, and the solenoid will get triggered to start refilling the tanks with freshwater until both are refilled.

Then, I get notified to add the salt again, and the whole thing repeats for eve and ever without a hitch.

Where am I going to fail on this? How can I improve on this? What problems do you see?

I know I just realized that the spout of water coming out of the Mag 9 needs to be higher than the top of Brute #2 or else I'll get water running out even if the Mag is off.

 

[smccart22]

I'll be tagging along, trying to work out a system for my upgrade whenever that comes lol, everything being completely automated with reliable fail safes would be a great thing, now if only there was a system that tested the water for everything we needed to an made the right adjustments lol, I'm sure sure there will be many with alot of experience in automating systems chiming in here, good luck

 

[jeremyjoslin]

I'll be tagging along, trying to work out a system for my upgrade whenever that comes lol, everything being completely automated with reliable fail safes would be a great thing, now if only there was a system that tested the water for everything we needed to an made the right adjustments lol, I'm sure sure there will be many with alot of experience in automating systems chiming in here, good luck

With the apex, I could eventually add in a salinity probe into Brute #2 to make sure that it's of adequate salinity before allowing to make the dump.

 

[mr86mister]

I'm not very familiar with the Apex systems Jeremy but with the salinity probe in Brute 2 could you get it stop the process if the salinity wasn't correct?

 

[pascal32]

This seems like a lot of effort

couple quick things:
(1) are you also doing topoff from the RODI bin? as drawn the bin serves no purpose once full, you might as well run the rodi line into the saltwater bin.
(2) if the RODI bin is used for topoff things get more complicated, easiest to remove exposure to TDS creep with a set of floats to refill once level gets to a preset point. this would require a pump to move water into the mixing bin - catch there is that depending on where you are between the two RODI floats you could be short on water - so you need a low float in the RODI bin to stop the pump from running dry - and a timer to come back and finish the job (genesis storm manages TDS creep)
(3) for the mixing bin can the apex set timers for over a week? I've never tried on the ReefKeepr, not sure it can.
(4) make sure you can remove 40 gallons from your sump (shut off return pump, heaters, skimmer, etc) - pump would likely need to go in the main area to get the volume
(5) what happens if you forgot to fill with salt and you replace with pure RODI? is there a toggle setting on a switch (required change from last time to indicate you put in salt? I know I can't program that on an RKE)
(6) you need a sump low protect mechanism in case the return pump ever runs dry.
(7) depending on how your ATO system works the water level in your sump will not be exactly the same all the time - as such you will have slight changes in salinity over time. My ATO is set to run for 30 seconds when triggered and is forced off for 9:30 to prevent cycling.
(7) you will want to use genesis float switch brackets for this http://www.genesisreefsystems.com/proddetail.php?prod=04-000018-01 replace the tube included with 1/2" PVC. You can see a picture of a similar "stack" I built at a friends house: http://reefcentral.com/forums/showpost.php?p=19902891&postcount=25


If you can program this on an RKE I might go trade in my reefkeeper stuff - I know I can't do this with an RKE. I've read quite a bit on the Apex and like a lot of what I've seen - this is pretty impressive for something with a "user front end on it" - if i had a controller that I could program myself directly (like Nate's) I wouldn't be concerned, but on these "simplified controllers" I'd make sure this is doable with a small proof of concept test.

If I was you i would add some manual intervention. maybe a panel which has a few buttons:
(1) fill mixing bin
(2) after salt mix press button - mix for 48 hours and send e-mail when done (if the Apex can do that)
(3) initiate water change

with that said stop over and look at my genesis system, i know it's not cheap, but it is hassle free

 

[pascal32]

With the apex, I could eventually add in a salinity probe into Brute #2 to make sure that it's of adequate salinity before allowing to make the dump.

what is the reliability on this probe? I know DA has had nothing but problems with them

 

[pascal32]

here is a more manual approach: http://www.reefcentral.com/forums/showthread.php?t=1335781&page=12

 

[der_wille_zur_macht]

Pascal took most of my comments.

It sounds like you're planning on running the heater 24/7? I'd probably try to program it so it only came on for a day, right before "water change day." Otherwise you're throwing money out the window. Also I'd assume you'd want the heater off when the container was empty unless it was below the inlet for the mag pump.

Also - the "empty" position floatswitch on the mixing bin - if this leaves any saltwater in the bin after each run, you're going to get an upwards creep in salinity - unless you're checking/adjusting manually each run. In other words, I don't think you'll be able to just dump in a fixed amount of salt each time. If there's saltwater in the bottom of the container when the freshwater starts to fill it back up, you're essentially going to need less salt the next time. And since the saltwater left in the next run would be slightly higher salinity (assuming you hadn't corrected for it manually) you'll need even slightly less salt the next time, and so on.

In my own attempts to automate water changes lately, I had to do some soul searching. What exactly is it about water changes that I don't like so much that I want to automate them? Can I automate JUST those parts of the process without introducing failure points by automating "too much?"

Basically what I'm getting at is something like what Pascal is suggesting. Instead of letting float switches and controllers do everything, put a button panel there, and push buttons to trigger the critical steps. By hard-plumbing everything, you're saving the hassle of manually mixing water, carrying around buckets and hoses, etc - but you're preventing failure by letting a human do a gut-check each time...

 

[jeremyjoslin]

Thanks for the comments. I'm going to need to re-read everything here a few times to absorb it all. A few quick things though:
- yes, I guess the first bin is technically useless, except that I thought it would be nice to have a supply of freshwater handing if I ever needed to do an extra "emergency" change. Is this worthwhile?
- I was planning to place a T in my RODI output.... one to go to my ATO line which uses a series of float switches and electric solenoid that is only "allowed" to run for 15 minutes 4 times a day (via Apex programming).
- I also read Los' thread yesterday, but admittedly got a little lost in the details... my concentration was taxed by the end.
- I'm new to the Apex, and will need some help on the programming end.
- Soul-searching? Check. My work schedule is intense and sporadic. I'll never change water unless I can automate it. I even considered paying for the service at one time. When I do think about it and have time to do it, I'm unprepared with the water and infrastructure to drain the right amount. I need to plan the whole thing to be turnkey and built in.
- The pieces for the Apex will run about $100. The Genesis is like 4x that, isn't it?

Back to the drawing board......

 

[pascal32]

yeah, genesis if 500 for the pro, 400 for the entry system. This does include pumps.

I would be surprised if you could do all that for $100 if you account for float switches, used ports on the breakout box, and ports on the apex power strip (all have a value) + pumps.

Like Nate said, take the parts out that are a "pain". I have three bins, one for RODI (fills itself), one for premix, and one for mixed water. to move between the bins I have a labeled wall switch and a pump, no buckets. I do have a third switch to sump into the tank in case I want to (siphoning etc). if you added a fourth to dump the sump you could do everything with wall switches. just exposing you to a complete non-controler alternative view.

 

[insomniac2k2]

Keep in mind that if your planning on running a NC solenoid, that they don't normally have a very good long run time before they overheat. It would take a very long time to fill your reservoir via rod I and gravity feed. You may want to pull water from your fresh water reservoir with a pump instead.

 

[jeremyjoslin]

Keep in mind that if your planning on running a NC solenoid, that they don't normally have a very good long run time before they overheat. It would take a very long time to fill your reservoir via rod I and gravity feed. You may want to pull water from your fresh water reservoir with a pump instead.

Really? I use a solenoid (I bought from www.autotopoff.com) to ATO right now. I've been using it for over a year, and it is situated on the output side of my RO (so that the RO membrane has constant pressure and I get less TDS creep on first few minutes of use). It runs for 15 min at a time now, but I've used it to fill 100 gallons before without any issue I was aware of.

 

[jeremyjoslin]

yeah, genesis if 500 for the pro, 400 for the entry system. This does include pumps.

I would be surprised if you could do all that for $100 if you account for float switches, used ports on the breakout box, and ports on the apex power strip (all have a value) + pumps.

Like Nate said, take the parts out that are a "pain". I have three bins, one for RODI (fills itself), one for premix, and one for mixed water. to move between the bins I have a labeled wall switch and a pump, no buckets. I do have a third switch to sump into the tank in case I want to (siphoning etc). if you added a fourth to dump the sump you could do everything with wall switches. just exposing you to a complete non-controler alternative view.

Pascal,

Your advice is well-received. I'm going to re-plan with one brute, and more manual steps. Probably use ATO via APEX for FW fill up, and then manually start the exchange. Also, probably use float switches to guide efflux from sump as well.

You're right about the cost of Apex outlets... forgot that part. I did buy a bag of 10 switches though for like $35. Remembering to make any single float switch mission-critical, it will be nice to use them as guides and triggers during a supervised operation.

 

[pascal32]

sounds like you are on the right path

 

[Seefour]

Holy Crap! I thought my RODI system and ATO was complicated...

 

[jeremyjoslin]

So, I've gone back to the drawing board after taking in all the advice and comments. Thanks.

Here's where I am now:

1.) Water from RO/DI fills Brute #1 until float switch is activated and solenoid turns water off.

2.) I get an email from my Apex that FSW is ready to be made.

3.) I walk down to the basement and dump in a bag of salt and turn on the heater and mixing PH in the Brute.

4.) The next day (or in 2 days if needed) I can walk back to the basement and initiate a manual exchange.

5.) I turn a valve on the system's main pump manifold to start expelling water into Brute #2 which is now used to precisely measure the correct amount of waste water. A float switch in here will help me someone.... not sure, but for now I'll just watch the water fill the Brute since it should be quick.

6.) The contents of Brute #1 (FSW) will then get pumped into main sump by manual trigger of Mag 9.

7.) Waste water in Brute #2 gets dumped by opening ball valve at bottom which goes to floor drain.

This process will take more manual steps, but the entire process can be staged ahead of time and the manual process should take no more than 5 minutes.

 

[fpv930]

What will you be doing for an ATO? Or, do you already have it in place?

 

[jeremyjoslin]

What will you be doing for an ATO? Or, do you already have it in place?

Already in place.

I'm adding a T in my RODI output.... one to go to my ATO line which uses a series of float switches and electric solenoid that is only "allowed" to run for 15 minutes 4 times a day (via Apex programming). The other line off the T will go to fill the Brute. Each fork has its own solenoid valve.

 

[fpv930]

I assumed with the Apex... you had a system already in place. Looks like the design should work very well. I just changed water in my daughters 10g and my 36g bow. It really stinks carrying buckets. lol

 

[jeremyjoslin]

Sub-thought:

If I change the Brute #2 to rubbermaid stock tank (the bulletproof cattle water watering ones), I could use it for acclimatizing new fish after a separate QT process.