Sand

I do not like sand as the culprit because that would affect all stages of the pump.

By the way everyone, I did not post my theory first because I did not want to affect anyone's thinking.  If someone has a theory I never thought of then my opinion stated first might prevent them from posting.

Richard Neff
Irrigation Craft

Right Pumpdesigner, it is not sand.  "iron worms?"

I would have to lean towards pump runout to the right of bep, causing suction cavitation in the lower stages (that is if the numbering system for the stages goes from top stage to lowest stage) which is overcome as head is added in the upper stages of the pump.

You have to love a picture like that.  I think the real problem is lack of engineering or no engineer responsible for the design.  

BobPE

Corrosion, erosion or some combination.  I'm sure that much is already obvious. I don't know why that pattern.

PumpDesigner,
My guess would be a combination of corrosion and erosion. It would be interesting to see what the impellers look like, since they probably are bronze. Also, the wear ring area (suction side) of the impeller and bowl should give some clues, especially if the pump has a stainless steel wear ring. A lack of wear or erosion on a stainless wear ring would tend to point the finger at a chemical rather than mechanical problem. Sand erosion should leave some indication of wear in the wear ring area, although the area being eroded is typical of the presence of sand.  
Five years of operation may not be that bad for this type of pump, it appears to be a small pump (under 8" bowl Dia.) and is probably running at 3500 RPM, which will result in high velocities and relatively rapid wear.
Good Luck,
rcrum

PUMPDESIGNER,

I've seen this fairly often in submersible "deep well" pumps.  It often occurs in pumps with larger bowls, for example 10" and larger.  People think that submersible pumps will not cavitate because the impeller is always submersed, but this is certainly not the case.  It is not unusual especially when operating at 3600 rpm, to have NPSHr of more than 50' with high flow submersibles in the larger sizes.

However, your holding something back.  You didn't show the condition of the 1st and 2nd stage impellers.  I'm assuming they are not as badly "iron wormed", or should it be "bronze wormed"?  Are there signs of cavitation and if there is, on what side of the impeller vane (high or low pressure)?

I would say this is the result of insufficient NPSHa.  I'd be very surprised if this is a 1750 rpm pump.

PUMPDESIGNER:

Good site.

Common since or cents:

Corrosion, erosion, electrolysis or abrasives will affect the whole pump with one exception.  If there are shims between the impellers on the shaft it would be possible that the wrong shim material was used. This would allow “downthrust” to occur at this point.  In your pictures I did not see any heat marks to support this.  The fact of a 5 year run also takes manufacturing errors out too.  Only one solution left:

Heavy duty termites!  

Can we all assume that each of the 5 stages are the same?  That is, the same size impellers/trim.

Internal picture shows the discharge end of the bowl, looks like discharge cavitation to me.  Can you determine if stages 1 & 2 impellers were properly set (wear rig gap to bowl)?  A pic of the #1 & #2 impeller would sure be of interest.

PUMPDESIGNER:

I'm not real up on irrigation systems.  I could be wrong, but I doubt this is corrosion due to only being the top two stages.  With irrigation pumps being feed from a lake do you very often have corrosive water (low ph) to pump?  I would have guessed it to be close to neutral in most cases.

One other question is do you have any way to see if the holes in the pump caused any casing damage due to fluid cutting from your holy pump?  The replacement pump may be pumping some sand, dirt etc due to holes in the casing.

BOBPI & Kawartha ALL:

I think we need an extermenator to check the well.  I still think it's termites!!!   

I would think the impellers to be in relatively good shape.  The cavitation is occuring on the discharge side, but is it really discharge induced???  I would think vapor cavities are formed somewhere in the suction side of the impeller and are collapsing on the discharge side as shown in the pics.

There is definately a problem because these pumps should last for 20 or more years, even at 3600 rpm...

Abbrasives or corrosion would affect the whole pump uniformly, not just center on one or two areas, so thats not it because the rest of the pump is in good shape....

What else do you know about it pumpdesigner, or are you working with what we know too?

BobPE

BOBPE:

Before PUMPDESIGNER can answer.....  I would have to believe that the impeller of the fourth stage as a min has some damage at least on the tip and the impeller of the fifth stage should have a lot of damage.  The reason I say this is because pressure is causing the bubbles to collapse.  Am I on the right track with this?

Thanks

PUMPDESIGNER

Come back Richard, you have us all worked up over this one!

PUMPDESIGNER, BOBPE All:

I would like to add one other possibility for the fifth (top) stage.  There could have been some pieces of metal from the fourth defuser (bowl) that caused some mechanical damage to the top stage (impeller and defuser.)  This may have contributed to part of the failure of the fifth impeller and/or defuser allowing erosion to play a part of the failure on the top stage.  Is there a good way to tell the difference?    

OK Richard, quit laughing and fess up!

PUMPDESIGNER,

I reviewed the pictures and by chance visited the "Florida Irrigation Society" homesite, to find an explanation of the pump damage in your picture.

Very interesting and enjoyed the trip with you.

KAwartha:

Thanks for your hint, I found the answer on their site too.  I am not happy with their answer as to the reason.  It was poor engineering and bad pump selection and lack of flow control.  Increasing NPSHa is not a viable solution for them to be mentioning at all since the pump was fitting the sump.  If the engineer had the opportunity to design the pump and the sump, then he would make a foolish choice in increasing the sump depth to increase the NPSHa since the cost would be prohibitive compared to the correct pump.

Makes you think though that there are a lot of non wqalified people out there wasting peoples money pretending to practice engineering!!!


BobPE

On same page:
http://www.fisstate.org/test51.htm

Put up photos of the impellers and one more volute photo.

Kawartha, I really was not hiding anything, I swear.  You remind me of a 19th century poem, "...But Father, he always suspicions me".

BobPE, Kawartha, d23, and the others you are sharp indeed.

Pump is 6", 3450 rpm, and probably was originally a 1750 rpm design adapted to 3450 rpm.  NPSHr is about 30 feet in the area that they were attempting to use the pump.

I believe (with the others) that severe suction cavitation was the cause.  As the bubbles were imploded in stages 1 & 2 they caused the damage.  By stage 3 all bubbles were gone.

The last photo I put up of the volute may be diagnostic.  Notice the pattern of damage on that photo.  The path of the bubbles can almost be envisioned by the pattern.  The bubbles can be pictured being forced along the volute guide vane, then as they slammed against the outside edge of the bowl they imploded and "ate" the iron.

I would like more knowledge.
Stage 1 impeller shows some pitting, but nothing that would indicate a problem like the stage 1 volute indicates.
Stage 5 impeller is smoother and seems to have no problem.

I guessed at the time I took these photos and studied them 1 year ago that brass or bronze impellers may be less vulnerable to cavitation damage than iron, perhaps because brass is more ductile?  Does anyone know about this?

BobPE - Send me some of those faces, I ain't got none.

Richard Neff
Irrigation Craft

BOBPE, Kawartha, PUMPDESIGNER, All:

I cheated!!!  We manufacture multi-stage pumps.  Been there done that!  

BOBPE FYI; From experience with this type of pump I would expect the termites ate the impellers some too.

I still do wonder if there is some way to tell the difference between cavitation and  broken pieces starting erosoinal wear especially after five years run.  With multi-stage pumps it seems that the cavitation starts at impeller "C" (cavition) and doesn't stop until the top stage.  The bubbles will collasp at a given pressure.  Does this mean that the top stage(s) are damaged due to abrasives (pump parts) and errosion?   Any ideas?  

BOBPE:

You and Kawartha nailed this!!!  I put the purple stars for you.  I also gave PUMPDESIGNER a star for making us think!

 

BobPE, you are exactly correct about the lack of engineers.
The reasons for those pumps is low price combined with a lack of knowledge, but then are not those sisters?

I took the photos about 1 year ago.  
The particular brand of pump is one of the lowest price pumps on the market.  I have been interested in their pumps.  But when I saw their curves and the NPSHr values on the pumps I blanched at the thought.  That manufacturer controls a large portion of the market.  And worse yet, as Kawartha said, many of their pumps require submergence of greater than 50 feet, and no one knows.  

As Kawartha correctly said, submerged is not protection against insufficient NPSHa.  But as long as no one is called to account for the problem?

Richard Neff
Irrigation Craft

pumpdesigner:

what was the suction entrance like?  I would assume the factory bell was on there with no other structure, this entrance cavitated too and damaged the first impeller but this was not the cause of the other cavitation as you noted.  Brass is extremely suseptible to cavitation, but from the pictures, there is no way cavitation was occuring inside the impellers.

I had another post on here some time ago that presented a new theory on cavitation damage.  As we know the common term is that cavitation forms bubbles.  In truth, cavitation forms cavities of absolutely nothing and a little fluid vapor, similar to the vacuum of space.  When these cavities implode, as the limit of the vacuum cavity volume approaches zero, temperatures can theoritacaly approach the temperature of the sun at a minutely small point.  This is considered to be the force that causes vaporization of the metal that the cavity collapses on.  In short, cavitation will distroy any metal or material.

This was a good exercise, thanks for the challenge....

BobPE

Kawartha
I gave no one stars yet, was too involved.  Will now do so however.

I saved this post entirely.
I appreciated all the comments and I really needed them.  I studied the photos and worked up potential models for what could have possibly caused the problem.  With your input I am now more confirmed in my opinion and that makes the knowledge more useful to me.

BobPE
I wrote the explanation on the website.  I brought up the increase in NPSHa because I wanted people to think about that possibility.  But you are correct, that was not an option.  Remember who my audience was though.  But then again, perhaps I should show a little passion about the failure of the owners to use knowledgeable engineers.

d23
Yes about the water.  Surface waters often have high nitrates and phosphorus from fertilizer run-off, much more than you might think.  BobPE is correct however a cast iron volute pump should by my experience last 20 years without a problem.  Cast iron has no problem handling surface water, but cast iron does not deal well with stupidity.

d23
There seems to be some confusion about the stages.
Stage 1 is the first stage in the flow path.  Stage 5 is the last stage in the flow path.  Stages 1 & 2 had all the damage.  Stages 3, 4, and 5 have no damage, none at all.

Richard Neff
Irrigation Craft

pumpdesigner:

I did not realize you wrote that whole thing.  yes considering your audience, you were right to bring that to you audiences attention.

I use a factor of safety when designing verticle turbine pumps for NPSHr suppled by the manufacturer.  These types of pumps all have high NPSHr curves no matter the manufacturer.  You note that the pump manufacturer did not readily advertise NPSHr only solidifies my opinion that the person that designed the pump was not an engineer since they did not know they needed the value.  In the USA, that is illegal for that person to have designed that if not a professional engineer.  I have sent many problems like this and their illegal designers to court with my reports unfortunately, it is a common problem.

take care

BobPE

I think cavitation would more strongly affect softer materials in the impeller versus the carbon steel of the volute. It is the volute and not the impellers that show the main ware. To me, it looks like you have failure points on the volutes of stages 1 and 2 because they are your most susceptible to corrosion at the higher pressures (and yes, perhaps additional cavitation). Looking at the latter stages, which have not yet failed but appear to be near failure, I would guess this to be more of a material selection issue. It appears that a high carbon, low allow steel has been used in an application where corrosion protection is paramount. I would be interested to know the chloride level of the water and the material makeup of the pump casing.

PUMPDESIGNER,

I gave you a star also.  You really got this thread moving, and agree completely with your analysis.

BobPE

As I mentioned in my previous post, this observation is not uncommon with larger submersibles.  It is generally the case also that the impeller does not exhibit serious signs of cavitation.  It is known that different materials will behave differently in cavitation.  There is no question in my mind that a good quality bronze impeller is much more resistant to cavitation than C.I.  S.S. and Ni.Al-Bz are far superior to bronze.  Duplex SS is somewhat better and  I believe that Titanium is used in some high speed impellers because of its superior resistance to cavitation.  I think one of the reasons for this is the strong bonds between the molecules, however I can't give you a scientific answer.  Another characteristic of the metal which probably affects its resistance to cavitation is the ability bend or flex without shear (this is why C.I. is so poor).

I think that over the years I have seen a list of metals vs resistance to cavitation, but don't remember where to locate it.  Maybe someone on this thread will have such info.

I would enjoy much seeing that list of materials relative to cavitation damage.

BobPE
What area of the country you work in?  I have a lot of people that I would send to you.  I help a lot of people but advise them constantly to hire an attorney and consulting engineer.  Because of my position I cannot really help them.  I hesitate a little to refer however because to be honest I do not know any engineers that are really qualified.  It would take a while for me to find one and then trust him so that I could refer.

ChasBean1
Stages 1 & 2 have the lowest pressure.  Water flows from 1 to 2 to 3, etc. with pressure increasing each stage.

ChasBean1
This is my best guess as to why I think brass/bronze would resist cavitation better than iron.  The damage is caused by a shock wave emanating from the collapsing bubble (pressures can reach 22,000 psi I read).  That shock wave passing over a rigid material like iron would tend to shatter the material.  Shock wave traveling over more ductile or resiliant material would not crack or shatter the material.

ChasBean1
Whatever damaged stages 1 & 2 was gone by stage 3.
There are not many things that could be there and then gone.
Stage 5 was perfect.  Unfortunately I did not chop out stages 3 & 4 so I cannot say they were perfect like stage 5.

PUMPDESIGNER

Not an expert by any means, but... I think cavitation by all means is the culprit.  The pressure experted by the early stage pumps reduces the suction head on the latter pumps, thus reducing the likelyhood of exceeding vapour pressure.  I think that wear probably did not occur on stage two until stage one pump's output pressure dropped below a certain level due to the reduction in the impeller surface area.  If this system was to continue, I would predict that all five stages would wear through the stages until pump performance dropped below a certain point and suction could not be achieved.

I repaired many a pump back when I serviced oil and gas wells and wearing at the outside of the impeller was common for us.  We dealt with high pressures, corrosive chemicals and high flow rates.  Chemical wear was always more severe where flow across an impeller is lesser, i.e inner impeller and wear plate/ring.

Like I said, not even close to an expert, but, I would go after your hydraulic system as it is engineered before I worried about how much I paid for the pump in this case.  Cavitation that bad will kill all metals right when you can't afford it.

Great inputs by all and I back BobPE in particular. Infact this thread created much interest in me that I wanted to give it a good thinking before I comment.After you guy's momentum within one day, I kept on thinking.

I have some ideas and appreciate any reply.

If erosion is the problem the thinning of the casing should be uniform. But I see more thinning near the puncture.

As the puncture appeared at specific intervals of the casing periphery, I am thinking the bubbles were generated at the suction side and were thrown towards the casing. (it is nice to imagine bubbles are equishared by the vanes). There is a possibility of serial blasts after first bubble collapses. So how they travel towards second stage and only to second stage if I am wrong?

One good logic by Bob about the pump operating near BEP at later stages. It convinced me to a great extent.

Togel
I agree with the engineering before worrying about cost.
Two things however brought on that comment:

1. Many other pumps would have worked properly in that application because they have much lower NPSHr values.  The pump used dominates the market and is used often because it is the least expensive.  

2. I have seen the installation, structure, piping, controls, etc.  Although I have much criticism of the design and installation, there is really nothing much more they could have done or even should have done as far as pump submergence.  There is no reason to increase submergence to 30 feet at a cost of 10 thousand dollars for a deeper wet well when there are many pumps that cost just a little more and would work properly with 10 feet of submergence.

Interesting thing here.  The pumps that have 60% less NPSHr values also have higher performance in terms of efficiency and pressure and flow.  There is no advantage to having high NPSHr values which is why many pump manufacturers work hard on getting NPSHr values down as low as possible.

PUMPDESIGNER

quark:

each stage adds pressure and the cavitation is overcome by the additional pressure in latter stages in this case.  I think if the flow were more to the right of bep then cavitation would progress towards the upper stages.  I still think the entire pump and all the stages are operating out of their cure as a whole and individually.  Vetricle turbines are difficult to diagnose because of their position in the water column.  Usually the first sign of problems is the motor suffers fom vibrational loading if the column will transmit it.  I think in this case, the holes in the casing reduced the head and flow and caused the owner to take a look....

BobPE

BobPE
You could be correct about operating to the right of BEP.
I was not monitoring the system before or after pump replacement.
What is it in the photos that leads you to that suspicion.

PUMPDESIGNER

Pumpdesigner - I Had brain flatulation (5 is the highest pressure)... Looking at the pics I thought corrosion looked pretty high in stage 5 also (like a pitting type corrosion, not just a general oxide layer). It could be the cavitation in the early stages causing little pits in the iron, causing an accelerated attack on the metal in those areas. Or it could in part be sand/particulate as stated by Automatic2 way back when. Exellent topic. -CB

Pumpdesigner:

I hail from Philadelphia....Plenty of bad pump designs and lots of work here!!!  If you ever need help by all means let me know....

BobPE

quark
You may be correct about the chemical causing a more even corrosion pattern.  The "serial blast" you mention got me thinking of cases of very severe cavitation, is that what you were thinking?  I have seen and read about cavitation progressing from small independent bubbles into what they call "super cavitation" where large cavities form, and also another phenomenae called "cloud cavitation".

BobPE
Send me your info (pump@shadow.net) and I will forward it to City of Bal Harbor (north side of Miami Beach).  They got a doosy going on there.  I get adrenalin rush thinking about you in there.  I cannot even express how funny that would be, and it would be easy to.  I am tempted to quit my job just to do that one myself.  Perhaps it is a long shot, but the kid from out of town cares nothing about the local engineers, and I'm sure you have seen that problem too.

PUMPDESIGNER

PD!

Two characteristics of cavitation are 1. Coalescing of small bubbles into big ones and 2. When one bubble explodes, it's shock causes further blast of other bubbles due to high localised pressures. Generally with damage of that magnitude I thought there was no chance of escaping from one stage to the other.

As I don't have much knowledge and experience with submerged pumps, I couldn't understand NPSHA causing cavitation. Is it only due to reduced suction or any other factor?

If it is due to reduced suction, when the pump operates at higher pressure at further stages the operating shifts to the left and the problem of cavitation should reduce. As the 3rd, 4th and 5th stages are in good condition, Bob might have pointed out this.

One more question from my side. When you opened the pump did you see black bubbles on the casing (in wet condition) which if you press oozes out water and have rusty material inside? I observed some black spots in the photo but can't make anything out of it.

There are is an incredible amount of fear and superstition around cavitation.  It is the bug-bear that keeps engineers from sleeping well at night.  There are some hard pieces of information though:

- Cavitation is the formation and subsequent collapse of vapor bubbles in the flow.  Even if the bubbles are at near perfect vacuum (call it 1 psia for computational simplicity), with 50 psia at the entrance to the volutes you are only talking 50 compression ratios when the bubble collapses.  Since cavitation is the most isentropic compression I've ever looked at, the isentropic compression equation would show a maximum isolated temperature of about 1000F - which is pretty hot, but since the bubble is pretty small the BTU's available to boil metal aren't enough for heat to be the culprit.
- No one ever solved a cavitation problem with material science.  PUMPDESIGNER, if you find the list of materials with a cavitation-resistance factor, I'll bet long odds that magnitude of difference is pretty small among metals that could reasonably be used for a pump impeller.  A few years ago I was having a major cavitation problem with a downhole jet pump (it turns out the suction piping was plugged, but it cost a lot to pull the pump so I tried metalurgy first).  When these pumps cavitate, it is always in the throat after the mixing chamber (much like your cavitation in the Volute, my theory is that the bubbles form in the low-pressure area, migrate with the flow, and collapse as the pressure increases).  I used 6 different metals (four stainless steels, Silcon Carbide, and Tungston Carbide) and every one of them failed within a couple of hours and after 24 hours the damage pattern was identical on each of the pumps.  The only solution to cavitation damage is preventing the bubbles from forming - raise the NPSHa above the NPSHr or lower the NPSHr below the NPSHa.  The US Navy has done more theoretical work in cavitation than everyone else put together (they didn't like either the noise or the damage on submarine propellors) and their solution was to improve the effeciency of the "air foil" shape of propellors to significantly lower the NPSHr.  They did a lot of really interesting metals work on the props as well, but that was primarily to improve the perfomance in sea-water corrosion modalities.
- The key to understanding cavitation is that noise that the submariners were trying to get rid of.  When the bubbles collapse, you have choked flow to fill the void (i.e., since the body of the fluid is at more than about twice the pressure inside the bubble, the fluid flowing into the bubble is traveling at MACH 1).  The shock wave (sonic boom) is very dense and very high velocity.  When it hits something, the force is transfered and since the wave is very small it can't deform the metal it hits so the irresistable force knocks some molecules from the surface of the immovable object.  A few billion of these impacts in a small area make a lot of noise and remove measurable mass.

I hope this helps get the boogy man from under our beds.

David

quark
zdas04 was correct, the bubbles implode not explode.  It is not possible to understand cavitation unless one first knows exactly what is occurring.

Concerning the last two posts.  Cavitation is not a phenomenae that we can easily use intuition on unless we understand that it involves a phase change of water from liquid to vapor, then back again.  That phase change introduces dramatic physics that we would not normally suspect are present.  The pressures inside the bubbles exceed 22,000 psi, and light is emitted because a plasma is created (all electrons disassociate from the nucleus).  These events create a shock wave, which we then hear and which causes the damage.

For confirmation do a study on Google.  We are fortunate in that the US and other governments have poured millions of dollars into studying the phenomenae because it has bearing on national defense (quiet propellers and high speed underwater missles), and ship propeller efficiency (money!).

PUMPDESIGNER