New surge device design, opinions wanted

[Foxy Brown]

I'm working a new type of surge device, just in the design stage now. Since I'm moving across the country in the next month or two and my tank has been torn down, it will be a while before I get to build it. Nonetheless, I just had to ask for people's thoughts while the idea was fresh. To be fair, it's actually a modification on the Borneman design, but since that wasn't exactly new at time (toilet tank) I feel OK calling this new too. The idea is to be bubble-free and silent. See the diagram below.

The drain is a bulkhead fitting, and a valve is created using a rubber ball which seats onto the top of the bulkhead which must be solid (denser than water). This is contained within a cage to keep it in place but allowing it to rise a short distance vertically.

The valve ball is attached to 2 floats, using fishing line. The first (lowest) float is chosen to allow positive buoyancy but not enough to overcome the pressure of a water column above it. The second float is the trigger for the surge and should provide enough buoyancy to lift the ball against pressure.

As the body of the device fills, the heavy rubber ball is seated into the bulkhead and seals it. The first float will begin to pull up on the ball once the level raises enough, but will not be able to pull it free due to water pressure. As it continues to fill, the second float will finally lift the ball free to the top of the cage and the water will start to drain. Now that there is no pressure on the ball holding it down, the smaller float is sufficient to keep things going until the level drops below it. Finally, as the level continues to drop below the smaller float, the ball will again drop within the cage onto the bulkhead and reset the cycle.

The design is a compromise between cost and performance. The body could be made out of a rubbermaid and the cage out of bent wire epoxied to the bottom. However, by making it out of cut acrylic sheet, it prevents the formation of a vortex and the entrainment of air into the system. Doing this eliminates the famous flushing sound, and by also keeping the water level always above the drain, there should be no reason to have bubbles enter your tank.

There is also to possibility of tuning the setup. Volume per surge is controlled by the length of line between the 2 floats, and buoyancy can be controlled by adding fishing weights to the line below the first float. The inner edge of the bulkhead will probably need to be sanded a bit to get a good seal, but that's easy. Addition of the classic P-trap (S-bend) should prevent any rebound and oscillation of the valve from water-hammer should it prove to be problem, though I would expect the negative buoyancy of the ball to be sufficient.

So, my question is, does anyone have any thoughts or suggestions? I'd love to hear of someone giving this a try. If you all think it's a good design, I'll plan on including it in my next tank.

 

[MrRyanT]

I think its a good design with one exception. I wouldn't use fishing weights because I think they are made out of lead and I wouldn't want that in my tank. I'd really love to see one these action though, it really looks like it would work quite well.

 

[Foxy Brown]

Yeah, I thought about the lead issue. There are a few Eco-friendly varieties that use tungsten or bismuth which would be better. Tying on almost anything dense would really be fine too. The best idea would be to size the float right in the first place. There are a thousand different fishing bobbers and all of them CHEAP, so you could try a few.

 

[shroutk3]

I'm not sure why two floats are necessary. The top float will be on top of the fishing float for much of the time anyways, possibly tangling the line, but maybe I'm missing something.

Tim

 

[Foxy Brown]

If you used only 1, the valve would close again as soon as the water level droped just a little and surge would be quite short. Toilet flapper valves are buoyant on thier own for this reason, but they tend to stay open too long and let air in or cause other problems. The point here was to let you control buoyancy, volume, etc. seperately by changing the parts. You could also add another section of line to top of the upper float to suspend it from above when the water level is low and limit it's travel to only an inch or so. That would keep the two from bumping into eachother.

 

[dogstar74]

You could use a toilet flapper, so long as instead of fishing line, you used a solid rod. That way the smaller float would "push" down the flapper before the entire water was drained. Then use a floating plastic chain for the upper float. It could be whatever length you wanted.

Just my .02
Aaron

 

[BeanAnimal]

The ball will rise and fall with the top float. There is no change in buoyancy throughout the cycle, other than the depth of water on the lower float, which is somewhat insignificant. There is no mechanism to pop the ball off, as the water rises to the top float the top float will slowly lift the ball and let water drain, this will either suck the ball back down or allow a balance to be found (depending on the pump size).

The float in a toilet has controls the fill valve, not the flapper. The flush handle acts as a lever to pull the float open far enough so that it pivots past its negative buoyancy point.

 

[AcroporAddict]

I guess it would work, but there are simpler designs that work well without any moving parts. I designed and built one in 1995 based on the Carleson design. I won't hijack the thread, but if you would like to know about mine, shoot me a PM.

Remember, the more moving parts you have the more things can go wrong.

 

[kcress]

I agree with Bean's assessment.

I think there is no impetuous to have the plug "pop off". You will just end up with a metered leak that completely consumes the pumped in water.

If you replaced the line with a spring then you might have the plug pop off and away until the water level drops a significant amount - the distance that the spring had contracted.

 

[Foxy Brown]

Thanks for the great comments guys, keep 'em coming.

First let me address your comment Acro. From everything I've seen of Carlson's, they are utterly reliable yes - but tend to to inject lots of bubbles because it is based on a siphon. That leads to salt spray which is just different maintenance issue. The reverse Carlson is better about this, but I want water going into the tank during a surge, not out. If you've figured out how to make a Carlson bubble-free I'd love to hear about it. The other problem with the Carlson is the minimum flow required to get them to fire properly... since I'm only looking to put about 100-200gph through the surge system, but I want fast discharge (requiring a big pipe), this will be tricky.

Then for Bean's comments: (By the way, I use your "dead silent overflow" on my tank and love it... great work!) You are absolutely correct about how a toilet works, but this not how a Borneman surge works, so it's clearly possible to get the "pop-up" with only a float to trigger it. What you're calling suction pulling it back down is actually drag (better to say blowing it down), and it's the shape of the valve (ball vs. flat plate) that changes things.

You are also right that buoyancy doesn't change with depth (water is incompressible, so depth doesn't matter). BUT the forces acting on the valve do change - depending on whether it is up or sealed down. When it is seated, the force pushing down on it is the pressure of the water column above it plus the suction of the water in the pipe below it (as long as it's not full of air). BUT when it is unseated and at the top of the cage, it is the drag force from the water flowing past it into the discharge tube that matters (there is not net pressure difference anymore).

To outline these forces, I have to give some details. You'll have to forgive me if I get a bit technical, I have an engineering PhD, and love equations. For my particular setup (120g) I was looking for a ~5g surge with a flow of 2000gph. I will be using a 2" dia. discharge tube, and the body will be a 12" cube placed about 1' above the top of the water line in the main tank. The pressure exerted by the water (in psi) = H/27.71 (H is the height of water column in inches), and downward force is given by F = P*A where A is the cross-sectional area of the pipe. Since H = 24" and the pipe dia. is 2", this gives a downward pressure on the ball of about 2.5 lbs! As for the viscous drag once water is flowing, the equation is a bit more complicated. If you calculate the drag coefficient (~7) for a ball this size (2.5") in water and given the linear flow rate (~1 in/s) in the body of the device, you get a drag force on the order of only 1-2 ounces.

So, once the ball is released, it will see about a 2lb decrease in the force pushing down on it... that's why it will pop-up, and why it will stay up even when the level drops and it's held only by the small float. Try it yourself with a bathtub, a ball and some string and you'll see what I mean clearly, but you need a pretty deep water column (couple of feet) to make it work. That's also why the flapper needs to swing to vertical so quickly... to reduce the drag effect.

kcress, I do like your spring idea as it would help to move the ball out of the high-flow region near the opening quickly and into the lower-moving water above it, which should help. I still maintain though, that once it's released the small float can keep it up, as long as it's chosen properly.

 

[BeanAnimal]

You are absolutely correct about how a toilet works, but this not how a Borneman surge works, so it's clearly possible to get the "pop-up" with only a float to trigger it.

The difference is that Eric's surge used a toilet flapper. The flapper pivots past its negative buoyancy point

In systems similar to what you are proposing, a reliable trigger is made up of a float attached to a lever arm so that the motion of the float is amplified significantly at the valve. This help move the valve quickly past the equllibrium point where drag against the valve = upward force on the float.

What you're calling suction pulling it back down is actually drag (better to say blowing it down), and it's the shape of the valve (ball vs. flat plate) that changes things.

We can look at it from either perspective Drag is caused by the water moving past the valve, but there is a significant pressure (Bernoulli's Law) difference at the throat of the discharge piping We tend to try to avoid the actual physics when possible and keep things general for those following along

You are also right that buoyancy doesn't change with depth (water is incompressible, so depth doesn't matter). BUT the forces acting on the valve do change - depending on whether it is up or sealed down. When it is seated, the force pushing down on it is the pressure of the water column above it plus the suction of the water in the pipe below it (as long as it's not full of air). BUT when it is unseated and at the top of the cage, it is the drag force from the water flowing past it into the discharge tube that matters (there is not net pressure difference anymore).

Yes... but the problem is still getting the valve to pop into the position you want it so that the forces do not balance

To outline these forces, I have to give some details. You'll have to forgive me if I get a bit technical, I have an engineering PhD, and love equations. For my particular setup (120g) I was looking for a ~5g surge with a flow of 2000gph. I will be using a 2" dia. discharge tube, and the body will be a 12" cube placed about 1' above the top of the water line in the main tank. The pressure exerted by the water (in psi) = H/27.71 (H is the height of water column in inches), and downward force is given by F = P*A where A is the cross-sectional area of the pipe. Since H = 24" and the pipe dia. is 2", this gives a downward pressure on the ball of about 2.5 lbs! As for the viscous drag once water is flowing, the equation is a bit more complicated. If you calculate the drag coefficient (~7) for a ball this size (2.5") in water and given the linear flow rate (~1 in/s) in the body of the device, you get a drag force on the order of only 1-2 ounces.

So, once the ball is released, it will see about a 2lb decrease in the force pushing down on it... that's why it will pop-up, and why it will stay up even when the level drops and it's held only by the small float. Try it yourself with a bathtub, a ball and some string and you'll see what I mean clearly, but you need a pretty deep water column (couple of feet) to make it work. That's also why the flapper needs to swing to vertical so quickly... to reduce the drag effect.

I too like physics and math While I fully agree with the idea, the problem I see is the scale and sizing the pump properly. What happens more often than not is that the pump does not provide volume fast enough to move the float fast enough to allow the pressure gradient to "pop" the flapper and instead equilibrium is balanced and you have a slight ajar float allowing water to leave as fast as it comes in.

In an ideal system, you will be able to find the proper balance of pump/sub-float/main float. Most folks will give up in frustration long before they end up with a reliable trigger. That said, I would love to be wrong and see your setup work and be easily adjustable.

One thing to note: You mentioned the force acting upon the valve. Assuming the valve closes with water still in the discharge pipe (no air), you will have at least 1' of head @2" dia. being brought to an abrupt halt! That water hammer is going to be significant and hard on the discharge plumbing. I would avoid 90's and would use #80 pipe and fittings. A sealed air riser may help to buffer some of the shock with each valve cycle.



Let us know how things progress.

 

[RocketEngineer]

Ok, While I feel this would work, there are a number of variables that will determine how well it works so lets simplify this:

1) the system works on differenetial pressure. When the ball is seated, the water under the ball and the water above the ball are both pulling the ball into the bulkhead. When the setup is initally started it will be just air under the ball.

2) The smaller, middle float is strong enough to lift the ball but isn't strong enough to overcome the differential pressure on the ball when its on the bulkhead. So, when this thing starts to fill, the length of the line between the ball and the middle float is critical because if it is too short the pressure difference won't be enough to resist the pull from the float and you will just get a small drain equal to what you are pumping in. Make it too long and the system won't produce enough surge.

3) So, the upper float provides the pull needed to pick the ball up off the bulkhead against the differential pressure. Since this one is on a longer lead, once it has triggered the ball to rise, the ball no longer requires that extra pull since the first float can hold the ball up.

This design uses the same concepts as a toilet flushing except that the flapper/ball is allowed to seat back onto the bulkhead before the tank drains. This is useful for saltwater aquariums because it means there isn't a rush of air before the surge causing microbubble problems. However, this design requires some tweeking and has several moving parts and potential failure modes. Adjusting the pull lines between the floats and the ball will determine the length of the surge. Adjusting how fast the water flows into the surge tank will determine how often the surge goes.

I would make one significan design change that I feel will allow this system to work better: Allow the ball to rise totally to the surface. You have the ball limited in how far it will move once the system is triggered but if you let the middle float go all the way to the surface you will remove the risk of too short a surge. Once you get it set up you can add a stop if you want but initally I would just use the length of line between the middle float and the ball to set how high the ball moves.

RocketEngineer

 

[Eldredge]

I am confident that it will work. I designed a very similar device using a flapper with the overflow plugged. I used plastic bottles as floats and it took a little trial and error to get them right. As you said, you can vary the size of the surge by changing the length of the line. I also tied the top float so that it only dropped a little in order to prevent tangles.

I also had to allow some water to leak past the flapper (I think I drilled a little hole, but I can't remember for sure) in order to compensate for some of the problems that Bean referred to. This also helped eliminate the water hammer - I didn't know about the P trap trick. It was also necessary to adjust the fill rate to an appropriate rate.

In the end, my device worked pretty well once I got the two floats and the amount of water leaking past the flapper and the fill rate balanced.

I think your design is superior and will work quite well. You might want to allow for some "leakage" past the ball to minimize shock and noise of the ball slamming shut. Once you get everything balanced, I think you will have a great surge device. In fact, I think I will use your ideas for my next tank. Good luck.

 

[Foxy Brown]

Great Stuff!

Bean: I think we're on the same page :beer:. I'm totally down with the float on a lever idea. Perhaps a 10g tank with the ball-valve at one end along with the hinge for the lever arm above, and the big float moving up and down at the other end. That would give some nice mechanical advantage. I also agree about the Bernoulli principle. The easiest way to beat that is to get the ball up and away from the throat: 1-2 inches seems enough just from paying around in the bathtub . Nice idea on the sealed air riser for water hammer - with that flow rate it will have a TON of momentum. They even make rubber bladder-based water hammer arrestors for home plumbing, I was worried about that as you'll notice from the first post.

Rocket: 1) I totally agree! I figure the first couple of shots will have to done by hand (hold the ball down, then pull it up when it's full) in order to clear the air out of the line. 2) I also totally agree. Fortunately, it's pretty easy to play with the length of the fishing line, but it WILL require experimentation. That's the fun part anyway! 3) Bingo. The more I think on it, the more I like the idea of unlimited vertical travel.

I specifically avoided putting dimensions on the diagram because I wasn't sure what kind of distances would be needed to get it all working properly. Also, since there are so many different setups, the dimensions will change based on size/flow preferences.

There's one more thing I didn't mention about my particular setup because I wanted to avoid confusion. My circulation pump is actually an air-operated "geyser" pump (http://www.airliftpump.com). I'm doing this because I have a refugium below the tank and I don't like impellers chewing up my copepods and such. It delivers a sudden 1/2 gal shot of water every 15 seconds or so. I'm hoping that the pulsed flow will help the triggering, but I'd prefer if it worked well with a constant flow too since most people just use centrifugal pumps.

I'll work on a new version with all of your comments included and see what the new opinion is.

 

[Foxy Brown]

Eldredge: Thanks for the vote of confidence! I remember reading an RC thread back in '05 where some leakage past the flapper was used just as you suggested to eliminate the water hammer problems. I can't seem to find it now though. I was thinking that you could just cut a small notch in the lip of the bulk head with a dremel to allow a trickle of water through to do the same thing. I'm really hoping to do without it, but it's always an option.

Unfortunately, due to my upcoming move, my wife has forbidden me from starting any new "projects". So, it'll be a while before I can try it myself

 

[HippieSmell]

Good luck with this. I've tried a few different designs based around the flapper style, and it was a never ending headache. I could get it to work for a day, then the flapper would either stay open or get stuck closed. There is a lot of abuse going on inside one of these, and something invariably fails, at least with the bubble-less style surges. Keep trying, but test, test, test, then test again before designing an entire tank around this.

 

[kcress]

Remember a toilet avoids the water hammer because they have that large standpipe that connects the space below the flapper to atmosphere, it also provides the safety overflow.




BTW You could use small surgical tubing for a saltwater-safe spring.

 

[Foxy Brown]

Here's the new version, with most of the comments included, except for the surgical tubing idea (good one kcress).

Unfortunately, the standpipe wont work for water hammer because this design is pressure operated, it has to be a sealed version - as per Bean.

Thanks all! Anyone else?

 

[Foxy Brown]

After looking at this, I think it would work better with the hinge at the other end... New version coming soon.

 

[Foxy Brown]

OK, here it is with the hinge turned around. Now a small increase in the water level will give a bigger motion on the rubber ball. That may help get it to pop-up more quickly once it's released (assuming you can get enough buoyancy force from float 2). Ball-guide system is turned 45 degrees to prevent interference with the swing arm. Float 2 is now made from a length of "pool noodle" with a PVC pipe down the center to act as the swing arm. That lets you slide it in and out to optimize the water-level to valve-lift ratio and the buoyancy (longer/shorter noodle section).

Any more thoughts?