Skimmer effluent measuring device

[karimwassef]

I'm on the road for two weeks and my skimmer goes a little nuts. It discharges 200gal in a week and it becomes a crisis.

I have the effluent running to a small two quart bucket in my wash basin in the garage. So it all runs down the drain.

When I'm home, I usually check the level in the bucket and pour it out every couple of days or a week (one quart a day is normal, two quarts is wet).

My first idea is a 3 gallon bucket with a level sensor and a flush release. This way, every two gallons, my Apex gets an "effluent full" indication and a U pipe flushes the contents down the drain.

This would give me a "frequency" measure of 2gal flushes and that can be my "warning" signal.

What do you think?

 

[karimwassef]

<a href="http://s1062.photobucket.com/user/karimwassef/media/1348F12E-96C0-4A4D-AEAF-7470DABE230B_zpsajua2ixa.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/1348F12E-96C0-4A4D-AEAF-7470DABE230B_zpsajua2ixa.jpg" border="0" alt=" photo 1348F12E-96C0-4A4D-AEAF-7470DABE230B_zpsajua2ixa.jpg"/></a>

 

[sleepydoc]

200 gal a week! How does your ATO keep up, and how do your saltwater tank from becoming a freshwater tank???!!! Since you said 1 qt/day later on, I assume the 200 gallons is a typo.

If it's sitting in your wash basin, you could simply put a uniseal in the bottom of the bucket with a solenoid-controlled valve and have the apex do as you say.

Before you said you had a controller, I was actually thinking of a simple Arduino-controlled mechanism that would control/count how much was discharged, but the Apex would probably be better. If the problem is that your skimmer periodically goes nuts and starts pumping out massive amounts of skimmate, you could set the apex to turn off the skimmer and send you an alert if the bucket fills up in less than a specified time period.

 

[karimwassef]

The 200gal is accurate. I have a 600gal system and the normal effluent is 1-2qts a day. I need to figure out what happened but it did go nuts and I lost a lot of saltwater. My salinity is still only 22 after a frenzied call to a friend who intervened and added a bag of salt.

The source is likely my monster skimmer but I won't know for another week when I get back.

So it sounds like the bucketized time method is the best?

I was hoping someone had a slick paddle wheel or other premade device to give me an output rate reading.

The Apex can't measure event frequency so the bucket x time method requires me to notice it and take action. This is complicated by the fact that the Apex won't graph out non-sensor data (even though it has it and logs it).

 

[Pife]

Sounds like you just need to get a large skimmate locker with a float valve to shut off your skimmer.

 

[rwb500]

running a skimmer the way you have it set up (with no overflow protection) is not usually done, and you found out why. Most people using the drain line use a skimmate locker with a float valve to shut the skimmer off when it's full. i've never heard of a way to protect against overflows that doesn't require periodic emptying of a skimmate locker or similar container. i suppose you could have the skimmate drain out of a small hole in the bucket, so if it goes nuts the bucket will fill and trigger a float switch. but the small hole could clog with skimmate.

 

[karimwassef]

I guess I can use a container that'll max out depending on my travel duration. So at 2 quarts a day x 10 day business trip, I can use a 5 gallon bucket to turn off the skimmer.

Has anyone set up a self-adjusting skimmer? This throttles the air inlet or water outlet to optimize the effluent? Sounds far fetched when I can't even measure the primary variable (effluent export flow rate) but I thought I'd ask.

Maybe a camera in the skimmer top that processes visual data based on calibrated signals of good and bad flow? (Even more far fetched).

 

[sleepydoc]

Skimmers are so touchy that I can't see a self-adjusting skimmer. Doesn't mean no one's done it though.

Even if an Apex can't measure frequency, you could figure out a maximum, like 2 quarts/day, then setup up a holding container with a float switch and a solenoid or other Apex-controlled valve. Have the Apex shut off the skimmer when the holding containter is full, but empty it on a timed basis independent of the float switch.

i.e. you want a max of 2 quarts/day. You have a 2 gallon container, so it should last 4 days. Apex empties it every 4 days no matter what. If there is less than 2 gallons you don't really care. If the skimmer goes crazy and fills it up in an hour, apex shuts down the skimmer until the holding container gets emptied 3 days later, then turns the skimmer back on. If it's still going crazy, Apex shuts it down again as soon as the holding container is full. It dosn't fix the problem, but prevents the salinity damage until you get back in town and can intervene.

Make sense?

 

[karimwassef]

Sure. How do you clear the bucket? Actuated valve?

 

[sleepydoc]

Yup.

Found these with a quick search; I'm sure there are plenty of others. You might want something a bit bigger than a ¼" tubing valve to make sure it doesn't get clogged with skillet gunk. You'd probably want to flush it out periodically as well, just to keep it clean.

I don't have an Apex, so I don't know the capabilities/limitations, but you need to make sure it's capable of driving the solenoid (voltage/current) as well.

http://www.bulkreefsupply.com/water-valve-tunze.html
http://autotopoff.com/solenoid/

Edit: just saw this thread - http://www.reefcentral.com/forums/showthread.php?t=2441420

 

[karimwassef]

I use a solenoid to control my Kalk input. It works pretty well.

But that's clean RODI water going into my kalk reactor's input. Skimmer gunk need a toilet siphon.

Maybe use a level sensor to trip the shutoff, but still use a purge U tube to empty the bucket (like a passive surge).

 

[sleepydoc]

You could do something like the trap on a toilet, but the problem with that is it's totally passive and wouldn't give you any way to prevent the over-skimming problem.

The tube I have hooked up to my skimmer is pretty small - probably about ⅜" ID, and it never gets plugged, even though I thought it would. If it's at the bottom of a 18" container, there will be a reasonable head of pressure on it, and usually the gunk is at the bottom, so there will be more pressure flushing it through with the thinner skimmate at the end. That may be enough to keep it clean. I haven't looked, but I'm sure they make ⅜" or ½" solenoid valves. I would think those would be large enough not to clog. Who knows, maybe the ¼" would too.

Edit - found this post (#16): http://www.reefcentral.com/forums/showthread.php?t=2116220 maybe try sending a PM to insomniac and see if he got his/hers working

 

[karimwassef]

You mean that the purge would automatically restart the skimmer because the float switch would be back in the "empty bucket" mode?

I can create a very long "hold state" in the Apex so the passive purge doesn't restart the skimmer:

1. Effluent fills bucket = float switch triggers "full bucket" mode.
2. The level where the switch is triggered also initiates the purge.
3. The switch turns off the skimmer and holds the off state for 7 days.
4. It also triggers a warning alert to my cell so I can override if I want.

The only tricky part is getting the purge level just right so it doesn't happen before the float switch engages. Or so late that the skimmer is turned off before the purge.

One solution is a little more programming: the float switch initiates two sequential states: 10 minutes forced skimmer ON, then 7 days of forced skimmer OFF.

Every control outlet is expensive in the Apex, but input sensors are cheap.

 

[sleepydoc]

"The only tricky part is getting the purge level just right so it doesn't happen before the float switch engages. Or so late that the skimmer is turned off before the purge."

This is exactly the issue. If you have the switch set to turn off the skimmer, then it never purges because it stops filling. If it purges before the switch trips the controller has no idea how many times the bucket fills.

Would it be possible to have the switch say 1" below the purge level and have the controller sense when it turns off again? So the bucket fills, trips the switch, but keeps filling until it purges. Then when the controller senses the switch opening again it turns the skimmer off for x hours/days.

Only problem with this is the switch getting gunked up and not working properly.

 

[karimwassef]

The two switch idea can work. I use this on my "leaky actuated surge" today.

The surge uses a DIY PVC and linear actuator valve to release the surge and then close again. It's an imperfect seal so it leaks some flow when off (not outside the pipe, down the pipe). So, if the time between surges is very long (8 hr quiet window at night), I need to restart the re-filling pump to keep the surge reservoir full.

Here's how it works: I have two switches, one near the top of the reservoir and one near the bottom. The refill pump pumps water into the reservoir until the top switch is activated. The switch being triggered turns the pump off. The reservoir leaks until it triggers the bottom switch, that turns the pump on again. This keeps the level in the reservoir between the two switches automatically.

In the case of the effluent bucket, I think one switch at the bottom will work. This switch will change state ( closed to open ) after a purge. That would shut down the skimmer and empty the bucket. So I'm capturing "purge events" instead of "bucket full events".

To turn the skimmer back on, I can manually change the switch state (to closed) mimicking a higher effluent level and have a latency programmed so it can't change state until the latency period. An hour should be enough for the effluent level to raise the switch to being closed naturally.

As far as gunk... I can use a standpipe with a mesh to allow just the clear liquid to indicate the level and trigger the switch.

 

[karimwassef]

Back to thinking about measurement... How about a micro circuit that measures the time between leading pulse edges and provides that duration reading. Dividing the bucket purge volume by the time is the rate.

There should be a simple micro circuit that can do that?

The next challenge will be using a data sensor on the Apex to show it. Maybe trick the temperature outlet to think its a temp probe. I have an unused temp outlet...

 

[e43devil]

http://www.maxbotix.com/SelectionGuide/outdoor-tank.htm#tank

 

[sleepydoc]

Back in my EE days I could've designed a high-low edge detector in my sleep, but things are pretty rusty for me now; I know you can do it with logic gates. Part of it depends on the capabilities of the Apex. You can also use and Arduino controller. That would allow much more complex calculations like rate of filling, average filling over xx hours, etc. You could then have the arduino signal the Apex, or control the skimmer directly.

Doughboy used these as a ATO level sensor - he just had it on the top and measured the distance with them. You can get knockoff versions from sainsmart or on eBay for even cheaper. These would require and arduino or some other controller

Edit - found this circuit. If the pulse is too short you could combine it with a latch circuit to hold it high.

 

[jamie1981]

Why not just use a peristaltic pump on the skimmer cup to pump out xxxml per day.
Have a overflow setup so that if the cup fills it just runs back into the sump?

I use a peristaltic pump on a timmer and have a float switch as a safety cutoff on my skimmer locker, if it gets full it shuts the skimmer down. and with the peristaltic pump and apex you can remotely turn it on to empty the skimtamite locker if you should choose to. This will also restart the skimmer and you will know how much you are running down the drain on top of it.

 

[karimwassef]

Great ideas guys!

So devil, your idea is to just use a level sensor and that captures volume over time. Little math gives me the rate. If I put it on a self purging tank, I can have a continuous volume vs time measure with a passive reset. Now, all I need is a way to get the level reading into my Apex.

Jamie, your idea is to sidestep measuring altogether and use a pump with a controlled export rate from the cup. Any more and the effluenr flows into a safety overflow back to the sump. This is highly automated and the overflow can have a warning trigger so I can take action. Great idea but I don't know if a peristaltic pump will work with gunky effluent? Any ideas on this?

Sleepydoc, since my Apex is networked, I can see what's hapenning when I travel. So getting it into that data stream is valuable. I'll post on the Apex forums for a way to convert level sensor readings to an equivalent temperature probe reading. Someone should have figured something like that out. Even if I don't use it for this, I've always wanted to measure my surge reservoir level.