Seawater Pump 10

[PUMPDESIGNER]

I am being asked to replace the pumps on a four lane underground tunnel that is below sea level.
Project is overseas.
The pumps dewater the tunnel which appears to have moderate leakage from the ocean a few feet away, why I do not know.

The pumps in place fail constantly and they are tired of flying the manufacturer in to repair and replace them.

I have pumped seawater before, but not enough to know for sure how to obtain the maximum service life.

My first approach is to provide a multi-stage turbine, closed impellers, submersible motor, all 316 stainless. We do not have to pump solids, we can install self cleaning intake strainers.

The original equipment we are replacing is a rail type system for sewage, open impellers, cast iron body, bronze impellers.

We lift the water 40 feet, then move it 300 feet horizontally to the ocean to dump it.

Each pump will move 300 gpm @ 80 feet TDH.

Anyone with a firm opinion about the most reliable method possible?

PUMPDESIGNER

 

[Kawartha]

Hi PUMPDESIGNER,

It's been some time since I've looked at Eng-Tips, and who do I find in the center of an interesting topic but PUMPDESIGNER.

This is an area that I spent a good deal of time with over the last few decades. I think that many have already indicated that this would probably be a good application for Ni-Al Bronze or Duplex SS. There are two major European Companies that have extensive experience in this field - KSB (Submersible Pumps located at Homburg, Germany) and Pleuger - a sub of Flowserve (Submersible Pumps located in Hamburg, Germany). Unfortunately dealing with the knowledgeable individuals at either of these companies is not so easy - they are located at the manufacturing facilities, not at the sales offices which you would be able to contact.

The material of choice about 10 years ago was Ni-Al Bz which replace Duplex SS, for offshore oil production platforms, fire and ballast pumps. A full chemical analysis would be necessary to make the best selection.

These pumps are generally driven by close coupled, water filled submersible motors. Reliability is generally extremely high and must be for offshore production applications. I think I saw in one of the responses to you that the motors are filled with distilled water. This is not correct, they are filled with clean potable water. One of the worst things that can be done to these motors is to fill them with distilled water - it is very corrosive. There are also oil filled motors available.

Had pretty good experience with small submersibles (up to about 40 hp) in Cast Iron bowls and Bronze impellers, in cold sea water, with limited contaminants. Sea water temperature is a major factor in deciding what material to use. As I recall the CI/Bz would last several years in cool sea water (< 25 DegC).

I still have some friends in the industry if you'd like me to try and find out the latest solutions possible, but I certainly would not be averse to recommending the use of submersible pump/motors, if the system design warrants this type of pump.

 

[PUMPDESIGNER]

Thank you Kastner for your comments. I have been leaning towards Duplex.

Hello Kawartha, you have been slacking off, but I am sure you will not be able to just sit around and do nothing. Your comments are good verification. I will need about 4-40 hp pumps, so if you have a model number or a contact name email it to me please, I would appreciate that.

I appreciate all the help, good help too.
I am waiting for a few companies to get back to me on the pumps.



PUMPDESIGNER

 

[KASTNER]

Hi Pumpdesigner

If you are from the US you should go in contact with:

*************************************
KSB Inc
4415 Sarellen Road
Richmond, VA 23113
Phone: 804-222-1818, Fax 804-226-6961
*************************************

http://www.KSB-INC.com

*************************************

There you will get all answers about the wanted submersible pumps.

Nice Greetings from Germany

 

[PUMPDESIGNER]

Thank you KASTNER

PUMPDESIGNER

 

[wellman]

Pumpdesigner

You may also want to try robbcopumps.com. We have several submersible pumps on oil plaforms in the Carribean. They are all AL Brz 958.

Thanks

 

[Artisi]

Hi PUMPDESIGNER
how are you getting on with this problem.
I would reinterate what I said before - the application is not all that arduous - it's been done over and over. I would suggest that it is more likely an application rather than a pump material problem.

before heading down the path of pump materials or even changing the pump manufacturer or pump style I would be looking closely at the application and the pump operating conditions, from my experience in the heavy end of the pump industry I would think it will be application rather than materials.




Naresuan University
Phitsanulok
Thailand

 

[PUMPDESIGNER]

Hi Artisi,
You are correct about the materials.
We can easily and cheaply do the job with 316 stainless.
We decided in the end that the pumps do not work continuously, perhaps a few minutes a day for exercise, plus operation for rain and some leakage.
Therefore, although warm saltwater can eat away the passivation layer on the stainless, that process would be so slow that it matters not. We think the 316 stainless is a little better than cast iron and bronze. (Not 316L, which would be preferable, but ordinary 316 because it is a standard material and affordable).

What do you think Artisi?

PUMPDESIGNER

 

[Artisi]

Hi Pumpdesigner
I think the materials are ok - is the failure being experienced a material failure or a mechanical failure.
I am not sure on this point.

Naresuan University
Phitsanulok
Thailand

 

[PUMPDESIGNER]

Excellent question.
I have analyzed what they have been doing.
They were sold 5 "Non-Clog" pumps (4-40 hp and 1-10hp).
The pumps have been a nightmare since they were installed. Constant bearing, seal, motor failures.
My analyses is that there is a severe radial thrust problem. The company that sold them the pumps now wants to sell them the same pumps agin, but this time in stainless steel. I told the customer in an email that purchasing the same pumps in stainless would be a huge mistake, same problem will occur, severe thrust loading and vibration.

That email got a big rise out of them, they had never told me (so what else is new), but they had been bothered by the vibration of the pumps when they ran ever since they were installed. They were considering how they could anchor the pumps down more fearing they would tear out their concrete anchors.

Sooo, they said they agree with me based on the vibration, and now they are waiting for me to offer the solution.

Those "Non-Clog" pumps are just huge single volute casings, big time radial thrust when off BEP. I suppose I could offer to force them to operate at BEP, but by now the pumps have been torn up pretty good, I say start over and put in good controls and multi-stage turbines, no radial thrust at any flow. Then I put a self cleaning rotating intake screen onto those turbines, will be there 20 years.

Any opinions?

PUMPDESIGNER

 

[Artisi]

Hi Pumpdesigner
Sounds to me like a typical problem installation.
What you are being told is the result of the problem not the cause.
I fully agree with you about not just simply changing the pumps for a different material, this doesn’t remove the problem and means further failure.

I don't have any problem with what are probably 'non-clog" single volute sewage type pumps, there are thousands of them running round the world in some pretty awful conditions. This project sounds like a typical application for a stnd. mass-produced rail mounted pump or free standing submersible pump unit - even a stnd. end suction centrifugal pump in fairly stnd. materials in a dry well situation, although more expensive for the construction and uses more real-estate.

Can I assume they will be Flygt, ABS or similar high quality manufactured pumps - not some "el-cheapo" copy.

If these pump units are from a reputable manufacturer, they will be designed to operate across the pump curve without much stress, especially as they are such small units. At this pump size, bearings and seals tend to be well over-sized and shouldn't give any problems.

Why do they vibrate - when do they vibrate ?? always a good question. I doubt it is actually the pump causing the problem - more likely unsupported pipe work or pipe work vibrating at blade pass frequency or at its natural frequency.

What is the method of installation, are they mounted in a sump, are they float level controlled, are they fitted with non-return valves so when they shut down the pipe line doesn’t drain back to the sump and start the pumps up again and again and again.

What is the failure mode of the seals and the motors. Are the pumps running dry, are the pumps running with little thru-put causing seal problems. Do the seals fail resulting in water ingress to the motors and then motor failure, you need to know.

As for multi stage turbines - for this application I think you are buying more troubles. As this is an underground roadway - I think it reasonable that the pumped liquor will contain a lot of fine abrasive material - unless there is a settling tank or similar you will see a considerable solids load at times as you will not be able to screen these fine solids.

Before putting my name to any concrete suggestions I would be looking for a lot more detail from the customer, if they can’t come up with the detail I would walk away from it. For me I wouldn't touch this job without a full engineering report from a qualified and trusted pump engineer or a site inspection by myself, I’m sure you don’t want any more sleepless nights.

Trust this rather quick overview is of assistance.


Naresuan University
Phitsanulok
Thailand

 

[PUMPDESIGNER]

Good comments Artisi, sounds like you have lots of experience to back up the knowledge that I know you have. Thanks for all your comments. You are not wasting your time with me, I am hungry always.

Existing pumps are described as follows:
Hydromatic/Aurora Model S8LX, 40hp, 14.25” x 13.25” Trim, 1150 RPM, Ns=3375
Wet Pit – pumps mounted on rails
Suction Specific Speed can only be guessed at, but the impeller is a 2 Vane and I suspect an Nss=~14000 or so.

The Nss 15000 is very high, and with only 2 Vanes this means that this pump would vibrate really bad at any flow rate even a little below or above BEP.
Then, the radial thrust is also huge, 2,450 lbs @ BEP, 6,324 lbs. @ 50% BEP, 7,602 lbs. at shutoff.
When you combine the radial thrust and the high Nss, and then run this pump at flows much below BEP (or even above BEP), you have some serious vibration by my estimation.

Do you know something I do not? I would love to learn something today besides how stupid I am.
I use radial thrust calculation methods of Cooper/Kittredge in Pump Handbook, (based on studies by A. Agostinelli, D. Nobles, and C.R. Mockridge), and also Stepanoff.

The specific link for this pump is:

http://www.hydromatic.com/nonclog/non-clog 1150.html

Would enjoy anything you have to say, I take criticism well I think, I'll ask the wife when I get home if that is true.

PUMPDESIGNER

 

[Artisi]

Are these units 2 vane impeller or recessed impeller, the curves seem to indicate recessed.

If 2 vane I can understand there could be a problem from my long experience with 2 bladed paper stock pumps - nothing what so ever wrong with the pumps and would run all day sweet and smooth in the test rig, but put them in the field with poorly supported pipe work - you could stand by and watch the pipe line destroy its self - especialy once everybody changed to light wall S.S. pipe.

Of course there wasn't any vibration until you started the pumps -therefore it follows that it must be the pumps fault - that was the story I was told one day. This was at 50 hz speeds so wouldn't expect any thing different at 60hz. I don't have any first hand experience with Hydromatis / Aurora but would suggest that they are producing pumps that are free from vibration when properly installed.
I would still be looking further into installation.

Naresuan University
Phitsanulok
Thailand

 

[Artisi]

As for asking the wife if you can take criticism reminds me of the English TV show "Rumpole of the Bailey" where Rumpole a Barrister, always refered to his wife as "she who must be obeyed".

Naresuan University
Phitsanulok
Thailand

 

[vanstoja]

Pump Designer
You posted your radial hydraulic thrust calculation corrections in the wrong thread. Your 740(bep), 1909(0.5QD), 2295(shutoff) pound values still seem too high based on the Pump Handbook equation FR=0.433KsHD2b2. The max K value at shutoff is about 0.36. I'm getting 426 at shutoff with that K and D2=14.25, H=80, s=1.0 (for pure water)and (by guess and trial) b2=4.2. What's the real b2 value? This is a static, directional load and seems unlikely to produce vibration at the equation force levels. The Agostinelli data is as good as any available and has been confirmed separately by CalTech engineers (Acosta,Brennan,etal) in 1985.

 

[PUMPDESIGNER]

Thank you vanstoja.
I'll check my work and get back to you.


PUMPDESIGNER

 

[PUMPDESIGNER]

Thank you very much vanstoja. I have made yet another mistake. My mistake was that I did not start with a clean template, was re-using a sheet with old entries and did not change all the data, then working on this project at night because I have not formally started working on the job yet, so I am working on that project late for fun when I am too tired to do my other work.
Fresh this morning, will start working on the job. Gee, this is embarrassing.
Here are the values based on Ns = 3400: (Max D2 Ns =2933, but I prefer Ns at trimmed diameter).
k=.433
Kr = 0.111 for BEP, .3170 for 50% BEP, and .3690 for 0 flow
Sp Gr = 1.0 (rainwater though seawater at other times.)
H(Head) is: 45@BEP, 63@50%, 87@0
D2=14.25
b2(Width)= 4.2 (good guess vanstoja)

That all works out by my calculations to:
129 lbs. @ BEP
518 lbs. @ 50%BEP
832 lbs. @ 0 flow

I have to now analyze this new lower thrust calculation, but any comments would be appreciated as to the importance of those figures.


PUMPDESIGNER

 

[vanstoja]

My calcs using your Kr values confirm your last posted RHT values for a maximum 14.25 inch OD impeller. Since it is trimmed to 13.25 you should be using that for D2 which gives 120, 481 and 774 lbs. for bep, 0.5QD and shutoff flows. You must have scaled the plot in the Pump Handbook to get three-decimal accuracy from the Kr curves. Correct?
If 3400 is your real design point specific speed then I calculate a design flowrate of 2638.7 GPM with the 45 ft. bep head (not the 300 GPM value you cited several times in the early posts). At this specific speed, Stepanoff's sigma equation 12.14 gives an NPSHR of 14.11 ft which results in a suction specific speed (SSS)of 8116 rather than the 14000-15000 you estimated previously. This is a moderate SSS value that would not suggest severe operating range limits. However, I hold data (possibly from Lazarkiewicz & Troskolanski)on high specific speed propeller pumps and turbines showing drastically increasing sigma values for decreasing blade numbers. For 2-bladed pumps and turbines the curve-fitted extrapolated sigmas are 3.0 for turbines and 3.92 for pumps. The latter value would mean an NPSHR of 176.4 ft. at rated flow if it applied to your seawater pumps. What are your actual NPSHR requirements at bep and other flow fractions? The question of cavitation-related damage is certainly unresolved as Artisi has pointedly noted. The damage to seals, bearings and integral(?) drivemotors suggests radial or axial overloading effects but this is not certainly cavitation-related or even RHT-related. What flow fractions are these pumps actually operating at? Are the seals being damaged radial(eg.,impeller wear rings) or axial(face seals)? What levels of axial hydraulic thrust loading are involved? What are the shaft strength capabilities? These are some of the design and operational questions I would seek first to answer and none of them thusfar involve material corrosion/erosion compatabilities.

 

[PUMPDESIGNER]

Impeller is listed as: 14.25" x 13.25".
However max diameter is 15.0", I assumed they taper trim.

Kr values I entered into worksheet are 4 decimal places.
Ok, I know, trying to obtain values off that little itty bitty graph they give you hardly justifies 4 places, but I scanned the graph carefully, blew it up big, took the data off it, then checked and adjusted my values against the examples they give assuming that they may have possessed the actual data.

Manufacturer says nothing on NPSHr or Nss.

I agree with your comments about more info. But then, I have never seen anything of yours that I disagreed with, not that I have taken the time to analyze many of your great detailed posts.

PUMPDESIGNER

 

[vanstoja]

Dynamic, rotating Radial Hydraulic Thrust(RHT)is a complex combination of hydraulic unbalance and fluid momentum changes in impeller channels, blade-vane interactions, seal gap and shroud gap flow effects for which I have not seen any generalized calculation methods thusfar. The best work in this area is experimental, mostly covering specific speeds in the 1140-1558(US units, RPM,GPM,ft.) range with one report for NSS=4213. The focus of experiments has been on dynamic stiffness coefficients for rotordynamics analysis and the dimensionless rotating forces don't always have the same reference as used by Stepanoff, Agostinelli and those reporting static RHT results so care must be exercised in interpreting the force levels involved. Force exitation frequencies include 1X, ZX (Impeller blade passing) and either random or discrete low frequency forcing functions due to rotating stall, etc. as low as
0.1X. Below are the recent sources I've seen on dynamic RHT forces in centrifugal pumps.
1.Kanki,Kawata&Kawatani,1981, ASME Paper 81-DET-71
2.Chamieh etal,1985,ASME/JFE V107, Sept., pp.307-315
3.Adkins&Brwennan, 1988,ASME/JFE V110,Mar., pp.20-24
4.Franz etal,1989,ASME/FED-Vol.81,July,pp.205-211
5.Bolleter etal,1989,ASME/FED-Vol.81,July,pp.187-193
6.Baun etal,2000,ASME/JFE,V122,Sept.,pp.598-605
Sources 2,3&4 are from Caltech using the same BJ pump and the same test facility. Source 6 is from UVirginia/ROMAC authors who have been recently pursuing pump hydraulic forces in parallel with Caltech. It is actually a static force paper dealing with optimum impeller offset in a circular volute pump. There may be several more recent German/Japanese papers that I have not seen.

 

[PUMPDESIGNER]

Thank you vanstoja, you read my mind, I was going to ask you for your list of references. I'll go find those references, big help, thank you.

The Caltech study especially looks interesting. I may be able to get some information from Goulds about a new series of pumps classified as "modified concentric", will pass it to you if I can get it from Goulds.

PUMPDESIGNER