Question about r.p.m. (1.500 vs. 2.900??) 2

[jacilore]

Hello all:

Regarding an installation I posted about in another thread of mine, (that's just a reference), I have collected 3 different offers for it.

The installation consists in an elevation from a deposit 300 mts. (985 ft.) high and 300 m3/h. (1320 us gallons/m)
Manometric height estimated in 340 mts. (1.115 ft.)

1) KSB MULTITEC A-125/3 10.1 (350 KW 300 m3/h 2.980 rpm)
2) RITZ 49200-11 (400 KW 330 m3/h 1.500 rpm)
3) PLEUGER 200 NM B (400 KW-330 m3/h 1.500 rpm)


From what I have read, the KSB meets all characteristics and requirements the installation needs, but the owners of the installation are used to 1.500 rpm pumps and are very skeptical about the reliability of the 2.980 ones.

Are they right to be so skeptical?. What can be in general the main withdraws for higher speed pumps?

NOTE: I wrote the main performances data in case I could be missing something, but I'm not requesting specific advice for the election itself. You are welcome for any suggestion, but my main concern is about the requirements or possible problems of the "fast" pump regarding the others.

Thanks and regards

 

[Artisi]

bradshi, good response - sums it up nicely. It comes down to using well designed pumps for the application, ie a unit designed to operate at 2 pole and not a unit that is marginal in design and over stressed at 2 pole.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[DubMac]

Yes I must admit this pump will certainly be required to be a 2-pole selection. The 1000 or so ft head is just too much for 4 pole, especially at 50Hz.

Jacilore, it looks like you will have to tell your customer that they are stuck with 2 pole due to head. See if KSB can even get a selection at 1500rpm, maybe not, it would be very many stages.

Must admit, I was thinking more along the lines of an axial split multistage Volute pump for this service. Just too much head here however to do that at 1500rpm. I stand corrected for even considering the 10 stage monstrosity that would require.

Continuing the discussion on 2 versus 4 pole in general however, and just for discussion's sake, I would have to stick to my guns on always recommending a slower pump if possible; especially in this service.

Bradshi, you obviously know your pump designs and your contentions on rotor stability are certainly correct...as long as we're talking about diffuser pumps in high energy services,i.e. barrel type diffuser pumps. Im having a hard time agreeing that faster is better in any way in terms of reliability.

The stiff-shafted rotors we've been building since the 80s are definitely better machines than than the spaghetti shafted 12 stagers used prior; and they CAN operate 24hrs a day for a year or two without need for substantial element rebuild. But we are now talking about an elevated class of turbo-machinery found in refinery charge pumps, decoking jet pumps, boiler feeds, etc.

Sold those pumps for years and they were always my favorites; truly the pinnacle of pump engineering. Speed does not kill in that world.

What is being offered to Jacilore here has nothing in common with that class of machinery however. This is for use in a municipal water delivery service, and I would say we are looking at an entirely different mentality. There is no Reliability Dept., and Maintenance Dept. is a guy with a truck and set of wrenches. Engineering is subbed out to consultants. It is required to buy low price, unless you have a strong consultant to hold spec. There will be very little monitoring eqpt., and what there is probably will never be looked at. That has been my experience with all except for the large-city municipalities.

Two pole speed pumps are almost non-existent in the Muni world; 4 pole is considered very fast. One of the reasons is that water, aside from being the most corrosive and destructive fluid we know of, typically contains sand fines that cannot be filtered out without great expense. Pumps in the muni world are designed to a large part with this in mind;.....THERE WILL BE SAND.

Now, I'm just remembering stories about the "pancake" pumps that we're looking at here (segmented ring section, if you will). I never sold them, but I sure replaced a lot of them with other styles. They can really be a Maintenance nightmare and they are parts monsters as well. If they are taken apart by anyone that is not a true Master Pump Mechanic; they almost never run right again; too many internal alignment issues.

Now I know KSB and Pleuger are both fine manufacturers (don't know the other one), but 3000rpm with sand fines in a ring section pump being maintained by government employees may be stretching the envelope too far.

Unfortunately, these Conditions of Service really limit the options. Hope there is no performance bond or extended warranty required for Jacilore's sake.

Please don't take my disagreement as argument for the sake of argument (I am, howver, a crabby old man); it is intended to: 1. help Jacilore with his decision, 2. further the discussion on these matters; the science is NOT settled, and 3. give me something to do this morning.

 

[bimr]

The decision to use either a single or multistage pump will be determined by the head the pump must produce to meet the required capacities. Some suppliers like to recommend a high speed small pump to be competitive, other suppliers might recommend a more expensive low speed large pump to lessen NPSH and wear problems.

Shaft speed is an important decision. Speed affects pump component wear as well as the pump size. High speed pumps cost less initially, but the maintenance costs are substantial.

If you double the speed of a pump, expect almost four times the shaft whip, wobble or run out and eight times the wear.

The wear rate of the components varies by the cube also

Example: At 1750 rpm. the impeller material is wearing at the rate of 0.020 inches per month. At 3500 rpm the rate would increase to: 0.020 " x 8 (23) or 0.160 inches per month. Likewise a decrease in speed would decrease the wear rate eight times as much.

The amount of shaft run out (deflection) varies by the square of the number

As an example : If you put a dial indicator on the shaft and noticed that the total run out at 1750 rpm. was 0.005 inches then at 3500 rpm the run out would be 0.005" x 4 (22), or 0.020 inches.

http://www.mcnallyinstitute.com/02-html/2-01.html

 

[GPRnD]

DubMac, no offence taken. Your experience and viewpoint have added a lot to the discussion.

As you point out part of the issue is the difficulty in finding a 1500 RPM. I ran the KSB online selection tool just for fun

(Note: this is in no way an endorsement of KSB over the other guys, I simply am more familiar with them as a competitor).

The program could not make a 1500 RPM selection as the head required was just too great. Perhaps the other manufacturers have more capability.

So it looks like you kind of need 2 pole. But if you expect sand as DubMac warns, then materials considerations are critical to long term reliability.

 

[bimr]

This comparison is between a Worthington NM 7 stage pump and a KSB MULTITEC 4 stage pump.

 

[GPRnD]

bimir, just to be clear, such comparisons of shaft runout are not useful in the OP's case since you are relying on the Lomakin effect to center the rotor.

You cannot extrapolate runout in the way you describe for flexible shaft machines. Such an extrapolation would only be valid for a stiff shaft machine such as an OH1, OH2 or BB1 style pump.

 

[bimr]

Flowserve, Worthington, and Pleuger are the same company with heritage brand names.

http://pdf.directindustry.com/pdf/flowserve/flowserve-pump-product-catalog/6194-57547-_44.html

The 150 NM pump actually looks like a better recommendation than the 200 NM pump.

 

[bimr]

The Multitec A 125/ 4-10.2 15.67 seems to be recommended by the KSB software.

 

[bimr]

Operation of a pump at 2900 RPM as compared to 1500 RPM has a much larger impact on the life cycle of a pump than the Lomakin effect does.

 

[DubMac]

Jacilore,

Sorry but I didn't even realize 2 of your selections are 1500rpm; you gotta go with the 1500 selection.

Bimr, what style pump is Flowserve's NM pump?? Pleuger ring section?? As an old Worthington guy, I would be interested to know its heritage.

Jacilore, would also be interested to know what the range of price difference is between the 2 and 4 pole selections. Obviously not asking for any confidential info, but just a approx. % range to go from 4 pole to 2 pole for purposes of this discussion.

If somehow you must use the 2 pole, do yourself a favor and make sure you make a statement on water purity required for service at that speed.

 

[GPRnD]

DubMac, a 7 stage 200NM weighs 1760 lbs. the KSB offering weighs about 930 lbs, and weight is a decent proxy for cost.

bimr, The NM selection you posted was for 1800 RPM. The OP needs 1500 RPM.

 

[jacilore]

Hi all again:

I'll need the weekend to read all the posts, unless it gets much crazier, (in the positive way).

However, I'll answer now the last easy question, for Dubmac: The KSB MULTITEC A-125/3 stage is about 27.000 €, the KSB MULTITEC A-125/4 stage is about 33.000 €, (different supplier), the RITZ 49200/11 stage is 80.000 €, and the Pleuger 200 NM (don't have the 150 NM offered) is about 90.000 €).

Regards and thanks to all again.

 

[miningman]

I cant speak to the electrical considerations but the head / flow characteristics are not excessive by underground mining standards. A 10 stage unit is not as obscene as one poster suggests.

On the basis that you get what you pay for, Pleuger are very well known in the mining industry and perhaps they are providing something that the alternatives are missing.

And if by chance the end duty location is an underground mine, are you ABSOLUTELY sure its clean water you're dealing with?? AFter 35 years underground , I dont think Ive ever once encountered truly clean water conditions

 

[bimr]

bradshsi,

The NM selection posted was for a 1800 RPM motor. A 1500 RPM motor will not have put out enough head for the application and will not work.

jacilore,

where did you get the 1500 RPM motor and pump selection? The online pump selector on Flowserve.com recommends a 1800 RPM motor.

DubMac,

The Flowserve website is calling the NM pump a Worthington brand. See the link.

 

[Artisi]

lets get back to basics, the OP has specified 50 hz operation either 1500 or 3000 rpm, if flowserve recommends 1800 rpm then the selection is wrong.



It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[stanier]

1500 rpm is 50Hz supply. UK,NZ, S Africe,Australia etc use 50Hz supply. Hence 3000 rpm rather than 3600rpm.

many water authorities decided that 1500 rpm was better becasue there was a eprception that things lasted longer. they have not done the numbers for decades. A true life cycle costing would see the reduced Capex for a 2 pole machine over a 4 pole machine. Also they tend to be more efficient but not always. They may have a steeper curve that gives better control and more suitable for parallel operation.

From the Opex size parts are smaller and thus less costly. you can replace a whole pump quickly.

Dubmac,
Concentrating of the initial capital cost of a pump is either ill informed or criminal. You need to do a life cycle analysis of the pump installation.

If you are worried about speed think about the Syndyne pumps that run at 26,000 rpm. Speed is all about selection of materials and quality pump products.

In the discussion on multistage pumps be careful to asses any change in operating conditions and how this may affect the balancuing disc. You want to avoid metal to metal contact.

Have you considered a Geho triplex diaphragm pump instead of a multi stage pump? Multistge pumps have fine clearances that can clog if there are suspended solids.

"Sharing knowledge is the way to immortality"
His Holiness the Dalai Lama.

http://waterhammer.hopout.com.au/

 

[GPRnD]

bimr, you can select a 200NMB with 7 stages which is allowed as a configuration. The FLS tool claims an efficiency of 80.9% for this setup with cast iron casings.

If you could not select this, you need to look at your selection settings. You may have something like API 610 construction set as a requirement.

As I noted previously it is important to specify HI class A for the test tolerances in order to get to true efficiency. I am not sure if the FLS tool is doing this correctly since it reports the same efficiency regardless of the test tolerance selected.

 

[DubMac]

Stanier,

No criminal activity going on here; possibly ill informed though. There was a time when I knew it all; however in my old age I've guess I've forgotten a whole lot of it.

Again, and mostly for Jacilore's benefit, we must keep in mind what level of machinery we are discussing. We are talking about municipal water service here; a part of our industry since Archimedes.

One of the most common themes in a muni consultant's spec is a speed limitation on the pumps. You'll rarely see the same in a refinery or chem plant service spec; 3600 is the most common speed you'll find there, and as you say, much, much faster.

I would say the speed concern has come from years of "life cycle cost analyses" rather than just perception. The Black&Veatch, CDM's, and Ch2MHills of the world are not incompetents; they have very good experience with pumps in water services.

Anytime there are solids present,(and when you're talking source water there WILL be abrasive solids); SPEED KILLS.

Not to scare Jacilore, but the ring section pump's reputation has never been too sterling as far as reliability. Just go look at the number of parts and pieces in those things; more crap stuffed in there than a Christmas turkey as they say.

Not saying there aren't tons of them around working just fine, but it doesn't take much to piss one of them off.

In thinking of Jacilore's decision, it might be criminal if I don't go down as saying; SLOW THE DAMN THING DOWN.

By the way, I know the Sundynes run just fine at 26,000rpm, and they are great pumps, but they are wearing out 8 times faster than if they were running 13000rpm, "and thats a fact Jack".

 

[stanier]

Hi Dubmac,

Great response. My point was only that things have changed and the consultants have not looked the new scenario. I know this is a generalisation but my opinion is they do not question but copy. Why their returns are based on using as few a number of hours for the fee. They are also under the pump to get the design out there. I have worked for the lot of them as a sub consultant.

You make the good point that pumps in refineries run at 2 pole speed. have you considered the costs involved in refineries wher eoil is $100 a barrel and petrol (their product) is $1.50 a litre.

Water on the other hand is $1 a kilolitre but has a lot of votes attached to its reliability of supply. Sewage is even worse as it has no value and even greater voter concerns about reliability.

n respect of wear BHRA published a study on slurry pipelining where the conclusions were that wear was proporional to velocity to the power 2.5 to 4.5. The variation comes from particle size and its distribution, particle shape and hardness, pH, temperature, angle of impact, pipe shapes, pump impeller and vortex design, material etc etc. Slowing the thing down wil reduce wear if solids are present. But the overall life cost may be greater as the bits cost more, the pump station is bigger, the VFD is larger requiring more cooling etc etc. the engineer has to do the numbers.

Water authorities are their own worst enemies. Thier tender documents rave on about life cycle costing then they accept the low cost tender with its poorly selected pumps. The hydraulic institute has some good resources on the subject.

"Sharing knowledge is the way to immortality"
His Holiness the Dalai Lama.

http://waterhammer.hopout.com.au/