iwaki and cavitation

[chimera]

very interesting.... i had just finished filling up a 20 litre bucket with RODI water and took it into the sump room. i wanted to top up the sump a little with RODI water but the bucket was fill/heavy and I slipped spilling a bit of water down the back of the sump. the iwaki pump came with a reasonably short plug so i used a small extension lead in the meantime to reach the powerpoint. the water got into the extension lead and iwaki plug socket and tripped the RCD, all power in the room went out. cool, so the RCD works at least. by this time the top inch or so of the main tank reverse syphoned back into the sump (as per normal) then stopped. i dried the plugs and flicked the RCD back on and all was back to normal.

so what you say? well, it got me thinking so I did some tests. i manually powered off the iwaki to simulate a power failure. as the water was back syphoning (and bubbles were coming out of the sump) i turned the power back on (this time, as the tank was reverse syphoning rather than when it had finished) now is my concern, the iwaki powered up but it was fighting against the air bubbles coming back down the pipe. i am guessing here that there were air bubbles trapped between the impellors - sort of like the effects of cavitation. to my concern, the iwaki was struggling against this and the water level in the sump neither went up nor down and the iwaki did not sound normal. in fact, you could hear it struggling. i switched it off then back on immediately, same problem. i switched it off and left it for about 20 seconds (until i saw a few more bubbles appear) then turned it on and it whirred up and was back to normal.

now this is a little concerning for me, if i go away for a long weekend and the power goes off and on again within say, 20 seconds or so (as the water is back syphoning, then it hits the top of the outlet of the sea-swirl and starts sucking bubbles down) then when the power comes back on, the iwaki is likely to do the same thing. it will struggle against pushing the air-bubbles back up and i would imagine eventually burnout the motor if left in this state too long.

i suggest anyone else with a similar setup to mine test this for themselves and post here, I'd be interested in your results.

Any thoughts on this? ideas to resolve? (other than a big-mutha UPS!)

[Ira]

How about an anti backflow valve before the inlet on the pump? Or leave it struggling for a little while, it should eventually clear itself, I'd think. Give it 5 minutes or something

[dan_from_nz]

can you put an s-bend in the pipping to avoid the air blocking getting down to the pump?

I.E. as it is sucking air in while still reverse syphoning if it goes down the tube and comes to a vertical s-bend (like under any sink) the air will naturally raise to the top of the s-bend while the water flows below, untill a ful syphon is created

[chimera]

not with a $600 pump, not willing to take that risk! it sat in limbo for about a minute, if it didnt sort itself out by then, i dont think it would at all. i didnt want to leave it any longer than that to test it as I could feel the pump casing starting to heat up (yes, even in under a minute!) i dont mind testing if anyone wants to loan me their one!!!

problem with anti-back flow valves is they restrict forward flow and also need cleaning regularly. for me, less maintenance = best!

the s-bend needs to be above the main tank water line. a possibility but more bends = less flow rate.

[dan_from_nz]

after writing that i don't know if it will work as it is air bubbles mixed in to the water and i think it would just flow through with the natural syphoning of the water, unless you can pressurise the system so the air does rise to the system

[chimera]

just edited my reply above re: s-bend. as i say, it must be above main tank water line otherwise it wont work. this will stop any reverse syphoning - you could even have a valve (similar to durso standpipe) at the top of the s-bend so air can suck in but no water can get out.

still not a viable solution! im surprised sea-swirl guys dont adopt a valve like the above into their units.

[Ira]

Hmmm, How about another pump force feeding the iwaki? If the Iwaki alone can't get the water moving so it can clear itself then two should be able to. If it'sa good sized pump you should end up with more flow too.

[chimera]

not possible and too complex (not to mention expensive). the pumps would have to be carefully rated to match and you'd never get it right! an interesting little dilema to say the least!

[dan_from_nz]

What about 1 90 degree bend with a join going straight up (like a stand pipe but instead of a 180 degree bend a 90 this wouldn't reduce flow too much and a pipe running straight up (to where you would have your air syphon hole) that way air would naturally rise and you wouldn't need 2 180 degree bends

[chimera]

Potentially, but the valve would be under immense pressure under normal operation (try over 3,500 litres per hour - not sure what that is in psi) so valve would have to be 100% reliable otherwise it pops off, water goes straight up and everywhere but in the tank!!!

I was thinking more along the lines of controlling this via electronics. Eg: having a device that the Iwaki hooks into that has a builtin timer delay so the power doesnt get fed back to the device on/off/on/off etc (thus has other benefits) Only problem with that is another area of potential failure to add to the puzzle.

Simple fact is I dont trust power companies around here! We've had about 8 power cuts in the last 6 months. Yes, thats RED BEACH, if anyone works for power company delivering to this region, you need your :evil: :evil: :evil: cut.

here's a photo of the back of the sea-swirl. it has a small 6mm or so hole drilled in the outlet into the tank (other side), just below the water line (which stops more water than is needed from back syphoning) water takes the path of least resistance so if it were done like a reverse-durso at the back, the water pressure would be greatest on the valve rather than heading around the 90 degree bend.

[wasp]

Dan is onto it, put it at the bottom of a u bend. This worked for me for a very similar problem, although not with an iwaki. What would happen is the impellor would cavitate in the air bubbles, but the lack of directional water movement in the pipe would allow the bubbles on the intake side to move up the pipe, and enough water would then get into the impellor to allow some pumping. This would then result in some more bubbles rushing down to the impellor with some getting spat out the other side, until the impelor was again cavitating. this whole thing would repeat several times for maybe 30 seconds or so until all bubbles were gone and things were normal.

Also, good attitude Chimera planning for this possibility. When designing this stuff and allowing for things such as 20 second power cuts that may or may never happen, it's best to say if it might, it will. That way you've got all possibilities catered for and can rest easy.

[dan_from_nz]

there are devices for slow power down and power up for electronics that hold large quantities os electric potential energy (don't ask me what they are it has been about 5 years scince i was told) our science teacher gave us a demonstration (accidentally) by unplugging somthing from the mains supply and getting a shock and explained how they had to power down, I think it had somthing to do with capasiters you could talk to an electrician, Or there must be somthing like a 60 second surge timer to stop the pump firing up for 60 seconds if there was a power outage or power surge but the problem with this would be constant slight micro surge could turn your pump on and off repetitavely

[dan_from_nz]

what about a small cheep power head in the return compartment of the sump pointing toward the outlet hole, this may just pressurise the compartment enough to squeze water into the impellor shaft removing the air blockage

[RnB]

how often power go off for say 20secs?

normally brown out, 2-3 seconds or minutes....

if minutes the water in the sump should keep the pump fully primed?????

cash going into your acc today for the iwaki 40 mate.. u see it tomorrow morning.

[chimera]

RnB - that is a very ignorant approach to the issue. If you can send me a letter signed by the Director of the powerboard saying if power does go off, it will either be a couple of seconds or an extended period of time and we will provide guarantee's on it, then I will be happy! :lol: I very much doubt this can be controlled. It IS a risk and an expensive one if the Iwaki blows up simply because of a power cut. There is ALWAYS a possibility. I like to cover my butt on all areas of failure - and thats just the attitude you tend to adopt when working in the I.T. industry!

Thanks for the dosh, post some pic's of the Iwaki in action!

[Pies]

If you are getting air trapped in your plumbing than you have an issue with your plumbing, fix the plumbing and the problem willbe resolved. MAKE sure the iwaki pump is well bellow the water level in the sump. You should be able to drill 'anti reverse syphon' holes in the return plumbing.

Now as for a 20sec of less power outage, the answer is obvious, not sure why you want to look for a more complex answerr, use a UPS. I have a UPS the runs my 2 Eheim return pumps ion my current system for about 25 minutes, its a 2000va and its batteries are fuxored. I suspect you could get away with an 800va, which are as cheap as chips (if not free). Check turners PC auctions, they often have millions of UPS at their auctions. I'can buy UPS battries for about $18ea.

Pie

[chimera]

i dont have an issue with my plumbing. the iwaki sits below the water level of the sump. the sea-swirl (at the other end) sits below the water level. when the power goes off, it reverse syphons (as the sea-swirl nozzles sits below the water surface and there is a small hole drilled in the seaswirl nozzle just above the output to reduce the amount of water reverse syphoning) As the water gets lower in the main tank, it starts sucking air back down the pipe - this is normal (and needed) as it eventually stops the water from reverse syphoning. my issue is when the power goes back on RIGHT AT THAT POINT IN TIME as air is being sucked back down the pipe.

UPS would be the go. I was keen on getting one but they're expensive new. Just waiting for one to show up at work... free or cheap!...

[Pies]

Seems strange that air would be sucked down the tube with the water and go as low as the pump?

Also a cavitation is when the pump is pushing more water than its sucking.

If the pump is bellow the water line, how can air remain trapped in the pump? Shouldn't the pressure of the water flood (thus priming) the pump and off it goes.

I had a look at you photos, I wounder if you need to have a bigger pipe taking water from the sump. Even perhapps the 'strainer' you have on the end of the inlet is restricting the flow of water to much? This is the only reason I can see for you having the problems your having, that the pump is unable to suck as much water as it needs. Open up the inlet to the pump or restrcit the outlet, and re-test. I am sure that will fix the problem anyway.

Pie

[suphew]

Check out the circuit for a power-on delay timer at this link, it's a 12 volt circuit but the relay can be used to switch 240 volts. also replacing the timing resistor with a variable resistor (pot) would alow you to change the delay.

http://ourworld.compuserve.com/homepage ... #therm.gif

[chimera]

Seems strange that air would be sucked down the tube with the water and go as low as the pump?

As I thought too, but it does and bubbles back into the tank. I thought the bubbles should go down as far as the water level in the sump but no further. I would imagine it's because of the shear height (assisted by gravity). When it starts sucking air it gets momentum and the air bubbles just keep going!

Also a cavitation is when the pump is pushing more water than its sucking.

Thats why I said "sort of like" the affects of cavitation. Cavitation is where the formation and collapse of low-pressure bubbles in liquids by means of mechanical forces, such as those resulting from rotation of a marine propeller, cause performance degredation. Good old dictionary.com

I had a look at you photos, I wounder if you need to have a bigger pipe taking water from the sump

The pipe going from the sump into the Iwaki is the same size as the inlet into the Iwaki. The restriction is the Iwaki inlet so it is getting enough water. Under normal operation, it works perfectly fine. As I say, its only when the power is switched on AS the bubbles are going down the pipe. Its possibly the fact that the return pipe is too small (reduced coming out of the Iwaki outlet) causing more immediate pressure and thus more tighter back pressure when it reverse syphons. The thing is it eventually needs a reducer anyway (needs reducing when it gets to the Sea-swirl) It would make sense to use a larger return pipe though so the flow-rate starts reducing at the main tank rather than immediately from the outlet of the Iwaki. Not sure if all this adds to the problem Im having here though. I think I'll fit a non-return valve (larger than the return pipe diameter - just have to use reducers - or bugger it, just enlarge the entire return pipe) to the return pipe thats the quickest and easiest way to resolve.

[Pies]

As I understand it, the inlet to the pump needs to be larger then the actualy pump inlet for the pump to work inrestricted. That is if the pump has a 25mm inlet, and 25mm outlet, you should use say a 40mm inlet pipe to the pump, so the pump can suck in a full amount of water and suction is only availabe as the pump inlet, not the inlet plumbing. If that makes sense. Thats what the pump people I have spoken to have told me, wide bore into the pump will keep water being PUSHED into the pump so it doesn't have to DRAG water into it, itself.

Pie

[chimera]

i understand the idea but dont agree with the theory. essentially, a 25mm inlet with a 25mm pipe feeding it is like a long 25mm inlet. IMO, this would only hold true if the water inlet is below the existing water level which is currently is (otherwise the pressure of the water to the inlet would need priming so it could be "sucked" into the pump). the only limiting factor i can see is a minor restriction due to friction in the pipe or bends leading to the inlet. i would think this is negligible over such a short distance.

i would say however, i agree in that the pipe 'cover/strainer' is more a limiting factor. i'll remove this (not really needed) and see how she performs.

i maybe wrong, but thats my theory!

[Pies]

Yeah I think your wrong (which I still think is why your having problems).

If you really think about it, it makes sense.

Real example. Ever seen a TUNZE Turbelle toploader pump or TUNZE Turbelle power head? If not have a look at the pics on their site. Now have a look at how big the 'strainer' is, why do they may it with such a big strainer? Its a pain having the great big drop down strainer, but they obviously manifcature them like that for a reason. I suspect for the same reason? Also notice the the bigger ehiem pumps have larger intakes than outlets?

If you have a 25mm inlet, ANY restriction on the inlet is going to cause problems, hence the reason to drive them with a much larger bore.

I got advice on this from Fluid Dynamics here in wellington, who make and use pumps for all sorts of comercial applications (LOTR films!). They were quite clear that for my closed loop pumps I needed a larger pump inlet than outlet. This is why my own tanks 'drilled bottom' changed from being 2x 20mm holes to 1x 20mm (out) and 1x 25mm (in). Stops cavitations and ensures the pump is being driven correctly.

I spent ages going through this in the planing and its why I changed my hole sizes, because I belive (through the advice I saught) that its important part of correct pump operation.

Pie

[chimera]

cool