Return Pump Cavitaion

[Salty Waffles]

Hello,

I'm working on a new setup. This is my first sumped tank.

I'm running a Mag12 return pump running internal. Not my first choice but I already had it. It may be getting bumped for a Fluval or Ehiem sooner or later.

I'm getting cavitation from the pump discharge. Looking to quiet it down. I have the pump laying on it's side with a 3/4" x 1 1/2" pvc bushing coming off the pump going directly to a 1 1/2" pvc 90*. (I realize the 90* fitting is less than ideal. I'll be changing this soon.) It's straight up from there through 1 1/2" pvc until it splits to two 3/4" pvc pipes.

I suspect the sudden jump from 3/4" to 1 1/2" is the root of the cause.

Turning down the ball valves on each leg of the return helps but I have to nearly shut off the returns to eliminate the cavitation.

Pointers? Should I gradually step up to 1 1/2"? Such as a 3/4" to 1" bushing into a 1" to 1 1/2" bushing?

In my line of work, we deal with large centrifugal water pumps. All of these are fitted with a cone shaped discharge pipe to take the 3" or 4" discharge up to 6", 8" 10" or whatever sized mainline. I've searched for something along these lines in pvc. No luck.

Open for suggestions, tips, etc...

[uncleof6]

On page two of the instructions for the Mag 12 and all Mag pumps 9.5 and larger, just below the performance chart, it is stated that 1.5" hose is the MINIMUM hose size that should be used to get maximum performance out of the pump. The pipe size "jump" is not the problem with these particular pumps. They are poorly designed considering to get any flow you need the 1.5" return line, when most every other would do fine with 1". Also, they are pond pumps, and would probably do better with much higher NPSHA, than you get with an aquarium.

http://www.dannermfg.com/Store/images/instructions/ZG100.pdf

This comes a a big surprise to many due to the relatively small diameter of the inlet and outlets on the pump volute. But as most should know, if they work with centrifugal pumps, that the size of the inlet and outlet do not determine the pipe size that should be used in a given situation/system. The actual job to be done, considering the NPSHR and NPSHA, determines the pipe size. Most hobbyists don't bother with the engineering aspects of the systems they put together, but they should.

Most would also know, that you almost have to go out of your way to get a submersible pump to cavitate, because there is zero fiction loss ahead of the inlet. (A flooded intake.) Things like an elbow on the intake, cavitate away... too shallow of a return section would reduce the NPSHA and could cause a problem as well as cause the pump to suck in air which could be mistaken as cavitation; and with a small return section, placing the pump too close to a wall in the return section would be seen as a restriction on the intake. Do not put anything on the pump inlet. Nothing.

Running this pump external, if the intake pipe size is not at least 1.5" or greater (2" would be best; assuming 1.5" outlet plumbing size. The actual size of the intake and outlet of the pump are irrelevant.) The pump would probably self destruct from cavitation, in a relatively short period of time. So running externally, with a 3/4" suction line, and a 1.5" outlet, your assessment would be spot on. But it does not really apply to a properly installed submersed centrifugal pump
.
From this, it can be seen that cavitation issues are caused on the inlet side of the pump, not the outlet. The pump is intended to run 1.5" outlet pipe. And yes, it makes a significant difference. Simple test for an NPSHA issue is to raise the water level in the return section, after making sure the pump inlet is free and clear—by several inches at the very least.

Another thing you need to do is pull the volute apart. Check for FOD, and damage. A sand blasted look to the impeller and volute, will pretty much tell you the assembly is lunched. Any other damage should be pretty obvious.

As for your return plumbing, and elbow off the pump outlet does not qualify for "less than ideal," and running it into another brick wall by splitting it into two 3/4" lines just kills the flow even more. Aside from the brick wall, 2 - 3/4" lines do not equal a 1.5" line. 1.5" return tubing or pipe means 1.5" all the way up and into the tank. Not much real benefit to running dual outlets anyway, in most cases. Since pressure is equal at all points in the return system, and the 3/4" pipe increases the pressure (smallest diameter) there is even less reason to suspect that the 1.5" pipe is causing a pressure drop sufficient to drive the pump into cavitation. You are pretty much negating the use of the 1.5" pipe. The only change you would see in the system is with velocity. Flow through the system would be constant, with the limiting factor being the smallest diameter.

What you could see on the return plumbing is a water tight, but not air tight join that would suck in air, and the amount would be directly proportional to the flow rate (reducing as the flow rate is throttled.) This is pretty common when tape is used rather than using a proper thread sealant.

[SFish]

On page two of the instructions for the Mag 12 and all Mag pumps 9.5 and larger, just below the performance chart, it is stated that 1.5" hose is the MINIMUM hose size that should be used to get maximum performance out of the pump. The pipe size "jump" is not the problem with these particular pumps. They are poorly designed considering to get any flow you need the 1.5" return line, when most every other would do fine with 1". Also, they are pond pumps, and would probably do better with much higher NPSHA, than you get with an aquarium.

http://www.dannermfg.com/Store/images/instructions/ZG100.pdf

This comes a a big surprise to many due to the relatively small diameter of the inlet and outlets on the pump volute. But as most should know, if they work with centrifugal pumps, that the size of the inlet and outlet do not determine the pipe size that should be used in a given situation/system. The actual job to be done, considering the NPSHR and NPSHA, determines the pipe size. Most hobbyists don't bother with the engineering aspects of the systems they put together, but they should.

Most would also know, that you almost have to go out of your way to get a submersible pump to cavitate, because there is zero fiction loss ahead of the inlet. (A flooded intake.) Things like an elbow on the intake, cavitate away... too shallow of a return section would reduce the NPSHA and could cause a problem as well as cause the pump to suck in air which could be mistaken as cavitation; and with a small return section, placing the pump too close to a wall in the return section would be seen as a restriction on the intake. Do not put anything on the pump inlet. Nothing.

Running this pump external, if the intake pipe size is not at least 1.5" or greater (2" would be best; assuming 1.5" outlet plumbing size. The actual size of the intake and outlet of the pump are irrelevant.) The pump would probably self destruct from cavitation, in a relatively short period of time. So running externally, with a 3/4" suction line, and a 1.5" outlet, your assessment would be spot on. But it does not really apply to a properly installed submersed centrifugal pump
.
From this, it can be seen that cavitation issues are caused on the inlet side of the pump, not the outlet. The pump is intended to run 1.5" outlet pipe. And yes, it makes a significant difference. Simple test for an NPSHA issue is to raise the water level in the return section, after making sure the pump inlet is free and clear—by several inches at the very least.

Another thing you need to do is pull the volute apart. Check for FOD, and damage. A sand blasted look to the impeller and volute, will pretty much tell you the assembly is lunched. Any other damage should be pretty obvious.

As for your return plumbing, and elbow off the pump outlet does not qualify for "less than ideal," and running it into another brick wall by splitting it into two 3/4" lines just kills the flow even more. Aside from the brick wall, 2 - 3/4" lines do not equal a 1.5" line. 1.5" return tubing or pipe means 1.5" all the way up and into the tank. Not much real benefit to running dual outlets anyway, in most cases. Since pressure is equal at all points in the return system, and the 3/4" pipe increases the pressure (smallest diameter) there is even less reason to suspect that the 1.5" pipe is causing a pressure drop sufficient to drive the pump into cavitation. You are pretty much negating the use of the 1.5" pipe. The only change you would see in the system is with velocity. Flow through the system would be constant, with the limiting factor being the smallest diameter.

What you could see on the return plumbing is a water tight, but not air tight join that would suck in air, and the amount would be directly proportional to the flow rate (reducing as the flow rate is throttled.) This is pretty common when tape is used rather than using a proper thread sealant.

Damn your the man

[Salty Waffles]

Thank you for your input.
On page two of the instructions for the Mag 12 and all Mag pumps 9.5 and larger, just below the performance chart, it is stated that 1.5" hose is the MINIMUM hose size that should be used to get maximum performance out of the pump..
Yes. Understood. This is the reason I went 1.5" out the pump.

This comes a a big surprise to many due to the relatively small diameter of the inlet and outlets on the pump volute. But as most should know, if they work with centrifugal pumps, that the size of the inlet and outlet do not determine the pipe size that should be used in a given situation/system. The actual job to be done, considering the NPSHR and NPSHA, determines the pipe size. Most hobbyists don't bother with the engineering aspects of the systems they put together, but they should..
Again, understood. I'm not an engineer, just a dumb farmer.

Most would also know, that you almost have to go out of your way to get a submersible pump to cavitate, because there is zero fiction loss ahead of the inlet. (A flooded intake.) Things like an elbow on the intake, cavitate away... too shallow of a return section would reduce the NPSHA and could cause a problem as well as cause the pump to suck in air which could be mistaken as cavitation; and with a small return section, placing the pump too close to a wall in the return section would be seen as a restriction on the intake. Do not put anything on the pump inlet. Nothing.
You understand my surprise then.
There is no plumbing of any sort on the pump inlet. I have a coarse strainer installed. Removing the strainer makes no difference.

I also believe the pump chamber to be adequate. It is effectively 20x12 with the pump inlet under 7" of water (to your point of NPSHR).

Running this pump external, if the intake pipe size is not at least 1.5" or greater (2" would be best; assuming 1.5" outlet plumbing size. The actual size of the intake and outlet of the pump are irrelevant.) The pump would probably self destruct from cavitation, in a relatively short period of time. So running externally, with a 3/4" suction line, and a 1.5" outlet, your assessment would be spot on. But it does not really apply to a properly installed submersed centrifugal pump
.
From this, it can be seen that cavitation issues are caused on the inlet side of the pump, not the outlet. The pump is intended to run 1.5" outlet pipe. And yes, it makes a significant difference. Simple test for an NPSHA issue is to raise the water level in the return section, after making sure the pump inlet is free and clear—by several inches at the very least.
Agreed on the point of cavitation usually happening on the inlet side.

Perhaps I have been using the wrong term all this time but when a pump is pushing too much water for whatever reason, you can hear the cavitation on the discharge side plumbing. Bubbles being exploded is what I understand to be happening here. That is what seems to be going on with my setup. Even with flow turned way down. Maybe 200 gph. I am quite certain the noise is coming from the pump outlet area.

Back to the point about NPSHR, in the course of experimenting with water levels to see what works, I can drop the level in the pump chamber to 1.5" above the pump inlet before it starts pulling any air, depending where the ball valves are set. Nominal running height is 7" above the inlet.

The pump inlet is centered in the 20x12 pump chamber.


Another thing you need to do is pull the volute apart. Check for FOD, and damage. A sand blasted look to the impeller and volute, will pretty much tell you the assembly is lunched. Any other damage should be pretty obvious..
This pump was new and never in water before this install. It's a leftover I've had from a setup that never ended up using it. But that is a good point. I'll pull the volute off and check for defects.

As for your return plumbing, and elbow off the pump outlet does not qualify for "less than ideal," and running it into another brick wall by splitting it into two 3/4" lines just kills the flow even more.
Can you clarify your point here please? Are you saying the elbow a bigger hinderance than I think? A major detriment?


Aside from the brick wall, 2 - 3/4" lines do not equal a 1.5" line. 1.5" return tubing or pipe means 1.5" all the way up and into the tank. Not much real benefit to running dual outlets anyway, in most cases. Since pressure is equal at all points in the return system, and the 3/4" pipe increases the pressure (smallest diameter) there is even less reason to suspect that the 1.5" pipe is causing a pressure drop sufficient to drive the pump into cavitation. You are pretty much negating the use of the 1.5" pipe. The only change you would see in the system is with velocity. Flow through the system would be constant, with the limiting factor being the smallest diameter.
I understand the cross sectional area of two 3/4" pipes does not equal a 1.5" pipe. IIRC two 1" pipes would get me in the ballpark though. Please correct me if I'm wrong.

The sacrifice in flow with the 3/4" wasn't a concern for me, nor was the elbow off the pump. I didn't plumb this tank with max flow as a priority. I have other means to move water in the DT and I wanted a slower flow sump. FYI

I initially planned to plumb this with two 1" returns off the 1.5" running from the pump. The two return bulkheads are 3/4". Hence the reason I decided to run 3/4" off the 1.5". Not to mention the easier packaging.

This is a dual overflow All Glass RR 125g tank. Two drains, two returns. It was bought 8 years ago before I knew of Herbie and Bean drains.

What you could see on the return plumbing is a water tight, but not air tight join that would suck in air, and the amount would be directly proportional to the flow rate (reducing as the flow rate is throttled.) This is pretty common when tape is used rather than using a proper thread sealant.
My only taped fitting is the 3/4" to 1.5" bushing on the pump. But this is underwater. Could it still be pulling air? I wouldn't think so. Everything else is a glued fitting with the exeption of the threaded nipples to attach hose to the bulkheads.

Another possible culprit could be the 1.5" union 10" above the pump I suppose. To my ear and experimenting, the noise seems to isolated to the pump discharge/elbow area. But its entirely possible the union is introducing air.

uncleof6, thank you for taking the time to respond. You have given me a couple things to think about.

If you have any further input on any of my questions/replies I would be appreciative.

[Salty Waffles]

Perhaps I have been using the wrong term all this time but when a pump is pushing too much water for whatever reason, you can hear the cavitation on the discharge side plumbing. Bubbles being exploded is what I understand to be happening here.
To clarify my understanding, excess velocity for a given size of pipe is a factor in the scenario described above.

[Salty Waffles]

To follow this up, I checked the impeller and volute for damage or defects. I filed some small casting burrs off the impeller but everything looked fine.

Because I had to know, I ran the pump without any plumbing attached. It makes the same noise. So either these Mag pumps are louder than I expected or I have a defective unit.

[uncleof6]

Yeah, the elbow off the pump is not a good thing at all. It is a brick wall. Some suggest two 45s rather than the 90, but the end result is the same. Only time 45s help is if you have two 90s, two 45s cut your loss in half. The larger brick wall is the tee at the split, and the reduction in cross-sectional area. In all cases a single return line of appropriate diameter, up over the top, will be the better performer.

Cavitation occurs in the pump volute itself. It is the formation and collapse of vapor pockets in flowing liquid in areas of low pressure (the output of the pump is higher pressure.) The audible alert sounds like sand in the volute, and the impeller and volute will have a "sandblasted" look, if left alone long enough. Usually there is a propeller (impeller) involved; Noise can occur with a sudden pressure drop such as across a venturi valve, or similar situation, but technically it is not really the same.

As I mentioned before, it is caused by excessive low pressure at the pump intake. E.G. the centrifugal pump is "sucking" in the fluid, rather than having a flooded intake. As I also mentioned, not usually an issue with submersible pumps. (unless the inlet is somehow restricted)

To clarify my understanding, excess velocity for a given size of pipe is a factor in the scenario described above.

Yes, this can be a factor. However, with 1.5" pipe, you will get up to around 81gpm (~4800gph) with minimal pressure loss and noise, assuming 20-100psi—not happening with a pump at 5psi, such as aquarium pumps; if @ 20psi one should be using a larger pipe size. Water hammer will probably become a problem (High pressure above 100psi) at around 126gpm (~7500gph.) These small pumps we use, can't even get close.

I understand the cross sectional area of two 3/4" pipes does not equal a 1.5" pipe. IIRC two 1" pipes would get me in the ballpark though. Please correct me if I'm wrong.

Close but no cigar... 2.25 1" pipes equal a single 1.5" pipe. Nitpicking, perhaps, but is what it is.

The sacrifice in flow with the 3/4" wasn't a concern for me, nor was the elbow off the pump. I didn't plumb this tank with max flow as a priority. I have other means to move water in the DT and I wanted a slower flow sump.

A common line of thought, but "intank flow" and "through flow" are two different animals, do different jobs, and are not complimentary or supplementary (they do not add together.) Power heads in the tank are an adjunctive aid to vertical mixing—the original purpose for their inclusion in systems that draw and return to the top of the tank. That is their only purpose.

Since an aquarium is NOT a single pass system, (RO for instance; skimmer) lower through flow decreases efficiency, whereas higher flow through increases efficiency. There is no method in the sump, in which a low flow is beneficial. Skimmer does its own thing, and is independent from flow through. Relative concentrations are affected by flow rate, and higher = higher. Another foreign concept that takes a while to get a hold of.

To follow this up, I checked the impeller and volute for damage or defects. I filed some small casting burrs off the impeller but everything looked fine.

Because I had to know, I ran the pump without any plumbing attached. It makes the same noise. So either these Mag pumps are louder than I expected or I have a defective unit.

If you still have problems without plumbing it is the pump that is the problem, not cavitation.

Mag drive pumps are noisy and hot, as well as inefficient (evidenced by the heat.) Not much else can be said. I have large 4-pole motors that are quieter. (noise is relative to perception and differs from person to person; noise is partially a function of RPM) My opinion is they (Mag Drives) should be retired from the hobby, along with most other re-purposed pond pumps. Newer pumps may cost a bit more, (some outrageously so) but in the long run, you come out ahead.