Well Pump Problem 1

[islandeng]

Hello all,

I am an engineer for a small island in the southern West Indies (about 140 miles north of Venezuela). We have been having a continual problem pumping brackish water to our watermaker.

The splines on the spline shaft shear off, allowing the motor to run at no-load speed until either the thermal protection kicks in or the motor burns out. This problem has happened three times, and the pump/motor combination has lasted between 3 months and 3 weeks. Our pump supplier is telling us it is an installation problem, however he also does not recognize the splines are shearing off…so I don’t trust him (meaning he only agknowledges the motor is burned out, even though the splines are obviously sheared).

The motor is a Grundfos (this has also happened with a Franklin motor) 2.00 hp, 60 Hz, MS 4000. The pump is rated for 25 GPM and 218 feet of head (according to the plate on the side of the pump. It is a 4 inch submersible model. The pump is made of 304 SS, as the pump supplier suggested it would work for our brackish water. However I feel the next pump should be made of 904L SS, to better handle the salt content.

The well is a 6 inch diameter 120 foot deep well with a 10 inch bullnose at the bottom. The pump must pull the water up the 120 vertical feet, then along a 1800 foot horizontal run through 2.5 inch PVC to an open atmosphere tank. As I stated before the water is brackish, I don’t have a salinity meter (yet) but I am currently assuming the salt content is about 75% that of seawater (could be more, could be less). The average temperature of the water is 89 deg. F.

I have spoken with the tech support at Grundfos, and they listed three potential problems for the repeating failure, frequent start-stop cycles (which we don’t have, as the pump constantly runs), pump overloading (possible, but the pump should be big enough), and misalignment between the pump and the motor (manufacturing defect occurring 3 times in a pump with an excellent reputation?). So my question after this lengthy problem statement:

1.) What is the cause of this problem? Is it simply the salinity content of the water, and we need to use the 904L ss? Could I drop a zinc down with the pump (we have a lot of boats here, so we have a lot of extra zincs)? Is the pump undersized for the application? I have been through the length of the pipe and there is no blockage.

2.) Are there any quick solutions? This is becoming more important, as our water supply is dropping and, well, that’s a bad thing. More importantly, long term solutions – is this a common problem with Grundfos MS4000? Should we look into other manufactures, or suppliers?

Thank you all for your help. I have I have given enough information.

 

[jay165]

Two problems jump out from your post:

1. Your motor rating is marginal for the service you describe. I calculate at rated conditions your required HP is 2.2. I'm assuming 70% pump and 90% motor efficiencies, respectively. I would have a motor rating of at least 4 HP.

2. 304 SS is the wrong material for brackish water. Bronze, CuNi, Duplex steel or even cast iron would be a better choice.

In addition, you may be overamping (the spline is acting as a mechanical fuse) your pump on start-up if you don't have a control valve on the discharge side. Even if you have a manual globe valve to partially close before starting, this would help in keeping the amps down as the system develops back pressure.

Hope this helps and good luck.

 

[BobPE]

Sounds like the pump may not be the problem, but rather the engineering behind selecting the pump should be looked into.

Obviously we need a lot more info, but from what you gave us, I can tell you need more help to solve this than just the pump supplier. Grundfos makes an excellent pump and to my knowledge, this model is not problem prone.

I am going to make an assumption here, and you know what they say about assumptions.....I am assuming that before all these problems, you had non-brackish water at your watermaker and things were fine. I am assuming that you needed a new pump for some reason and called a pump supplier to assist you. The pump supplier gave you a different pump that what was originally in service. At this point, I am assuming that you started to get brackish water in your watermaker.

Please let us know if my assumptions are on track and I will continue based on your input, we should all be able to help you here...

take care...

BobPE

 

[islandeng]

Jay,

I agree with you, i got a nominal requirement of 2.18 hp when i plugged in the numbers, but the tech support at Grundfos said that was well within their factor of safety for a 2 hp. And i also think the 904L ss is a better material choice.

Bob,

Close, but a little backward. Originally we were pumping (and still have the capability to pump) the water out of the harbor. So it was pure sea water coming into the water maker. This worked really well until people started anchoring around our pipe inlet, and then we started getting sand, waste, and other stuff into our strainers and filters, which not only shot our filter life (it got to one filter per week!) but it also reduced our water maker output. So we decided to dig some wells to avoid the anchoring problems.

The wells never pulled fresh water. They have consistantly pulled brackish water and have only been in service for a year or two. Our pump supplier (located in Trinidad) was on site when we dug the wells, and selected this pump then for us. We have been using this Grundfos MS4000 since the beginning and have been having problems for the last 7 months, where we have replaced the pump 3 times.

Following is some more information about the well, it may help:

The construction of the well is 6 inch diameter PVC pipe. It is 120 feet deep. Starting at the bullnose there is 10 foot section of periforated piping, then a 20 foot section of blank pipe, then a 20 foot section of periforated, then the rest is all blank pipe. The well lies in very sandy soil with some rock (it is called "blue bitch" here, but from what i can tell it is a weak, non-porus, volcanic rock) and some aggrogate (same bluebitch but 1/2" to 2" square chunks). There is no clay (relatively) and minimal organic material, and no trees within 30 feet of the well.

Thanks to the both of you for replying, if you need any more information, please ask.

-Loyal (EIT)

 

[BobPE]

islandeng...

thanks....assumptions are always dangerous, but for us here, they help us get conversations moving is directions hwere we can help people...

I see you treat brackish water...ok...

jay165's comments on material are very important in this application....it seems like you are aware of that...so if the pump supplier helped you with that decision, get another supplier...

Looking at your pumps system requirements, I, as did jay165, would assume you to have a discharge control valve on your system that allows you to regulate flow???? How do you know you are pumping 25 gpm? Also, how do you monitor water levels in the wells? Did you pump test the wells to determine hydraulic characteristics of what flow they can sustain? I get a dynamic head that is less than the BEP head you gave for the pump of 218 feet of head which means that if you are not regulating flow, you will run the pump off it's curve...

To shear a spline requires a pretty sizable force on a pump like this...the forces can be instantenaous, or cyclical....see if you can answer our additional questions....I think people will have some good ideas....

bobPE

 

[25362]

The estimated pump output would be:

205*25/3960 = 1.29 HP​

for an overall efficiency of 0.7*0.9=0.63

the motor would have to supply:

1.29/0.63=2.05 HP​

To j165: am I right ?

 

[islandeng]

Bob,

thanks for the questions, here are some answers that I've obtained through my own observations, or through speaking with our head engineer:

According to the head engineer, we do have a discharge control valve. However it controls the flow direction of the water, not the backpressure in the system (although it will contribute a little backpressure). The valve diverts the water from exiting into the tank to exiting back into the well. So basically we have a supply and return line on our system. This is to allow the pump to continue running regardless of our demand (or lack there of) and thus prevents cycling the pump. Generally the valve is held at 3/4 open, meaning that, in a perfect world, 3/4 of the flow would go to the tank and 1/4 will return to the well.

The data I gave, 25 gpm and 218 feet of head are stamped onto the pump itself, as maximum permissible (like an either or situation). We are actually pumping about 12.5 gpm. This is calculated by a clock and a 5 gallon bucket (timing how long it takes to fill the bucket and then doing the math). When this was done, the diverting valve was 100% open.

We moniter the water levels in the well with a string (seriously) that shows us where the water is. We also check for flow bi-hourly. I don't know if the wells were pump tested, They were installed and running before I was involved in this project. I would hope they were tested...but frequently "little" details are dropped out here.

I have no clue if the forces are cyclical or instantaneous. Judging from how the sheared spline looks, I would venture cyclical, but that is only a guess. I say that because i don't find chuncks of metal, and it doesn't look knurled up, it just looks worn down and fatigued.

Something else that I found interesting is that the motor's spline measures 3/4 of an inch in length. The pump's collar measures 1/2 inch in length. So the motor's shaft isn't obtaining the desired contact. I brought this to the attention of the support department at Grundfos, and they didn't comment on it (no, "we know, it is supposed to be like that" or "really? that's not correct, something is wrong there.") What did suprise them is that the motor's shaft is shearing, not the pump's collar. (To me that made sense from reasons stated above.)

Thanks for your help thus far. Again I will answer any follow up questions to the best of my ability.

Cheers,

Loyal

 

[jay165]

25362:

Where'd the 205 come from? - he said 218' in the original post. Also I used 1.02 for the sg of brackish water. Putting those two factors into your equation yields 2.22

 

[25362]

Correction: the estimate was indeed done on 218, not 205 gpm.

 

[25362]

To j165 you are right. I cannot explain (I wish I could)where did the 205 gpm came from. And you're again right in your last 2.22 HP estimation including efficiencies. How, then, did you arrive at 4 HP for the motor in your previous post ?

 

[jay165]

It's the next higher even number and I don't like odd numbers?

 

[islandeng]

I like your thinking jay165...

So i got a bit curious and actually measured everything out myself. This is what I found:

Diameter: 2.5 inches
Verticle Head: 130 feet
Other minor losses: 4.3 feet (length of pvc, two valves, one partially closed, and a couple 90 deg bends).
Desired Flow rate: 25 gpm
Specific Weight of sea water: 10.35 kN/m^3 (i know it's metric...)
Same efficiencies as Jay, cause they sound good.

From this I found that the required power is about 1.4 hp. A 2 hp pump should be able to do this, correct? The Factor of Safety is 1.44, but in this situation that should be acceptable... I'm looking for other, more experienced opinions than mine.

Jay165, I know you aren't a fan of odd numbers, but would a 3 hp be more appropriate, especially considering we have had this continual problem.

Bob, how about you, any new opinions from this data?

I want to recommend using a 3 hp 904L ss Grundfos pump, but I want to make sure that it will solve the problems we have been having.

Thanks to everyone for thier contributions.

 

[PUMPDESIGNER]

Hello BobPE, nice to see you around some.
Is there a check valve at the pump?
A check valve on top is good, but one at the pump cannot hurt. If the pump is a Grundfos, it has a check valve built in, like most other 4" itty bitty pumps. Is the check valve working?

If when the pump shuts off, the check valve(s) do not hold, there is a very hard fall of water back down the hole and some violent reaction at the pump and motor.

Regarding the horsepower and such. Most pump manufacturers don't ever hesitate to overload a motor, not a problem to them. You did not state what kind of submersible, but I would guess you have one of the formed stainless sheet metal pumps. You can get a heavier duty pump, with trimmable impellers, and a submersible motor with true roller thrust bearings. Lots more money though.

PUMPDESIGNER

 

[islandeng]

Pumpdesigner,

Nope, we don't have any check valves, other than the one on the Grundfos. However since the pump isn't cycling, that shouldn't be a problem...right? You are correct when i stated it is a formed stainless sheet metal pump. I like the suggestion of getting the heavier duty pump, however the extra money does present a slight problem...

Thanks for reply and the suggestions.

-Loyal

 

[d23]

islandeng,

You did not say how deep the liquid level is in the well. Circulating water back into the well can carry air to the pump intake. If you get a large bubble of air into the intake it is very possible to air lock the pump. If the motor continues to run under that condition the unit will over heat and cause a premature failure. Are there any indications of heat on your motor(s) or pump(s)?

A small low volume sub pump is most likely a radial flow impeller that has a very low Ns. In most cases I would expect the required HP to be a little lower by producing at a lower flow rate rather than at BEP. You may be better off applying a little backpressure rather than circulating fluid.

Circulating fluid back into a well is not a good design. Adding additional oxygen to brackish water could lead to scale. Have you seen any scale on the pump or motor?

On submergible pumps the power supply is very important. If there is something like an auto-re-closer on the power supply that could explain broke shafts. If the power blinks for X ms without opening the motor starter it could cause catastrophic torque loading.

Hope this helps

D23

 

[islandeng]

D23,

Thanks for the comments.

The water level is 15 feet below the top of the well, which means the water in the well is 105 feet deep. The pump is physically located 118 feet below the top of the well.

There are no indications of heat on the motor, which is why I don't believe our supplier when he blames the failure on the motor burning out. In fact in the last spline failure, it was caught quickly enough that the motor still checks out ok.

Not to sound ignorant, what do you mean by "scale" on the pump or motor. When we pulled the pump and motor out of it was pretty clean with no signs of wear, except fot the splines.

Power supply is the only thing i have not yet checked. That is one of my projects for tomorrow, so I will let you know what I find out. Thanks again for your help.

Thank all of you for your help.

-Loyal

 

[PUMPDESIGNER]

d23 had a good idea about power.
Just a simple loose wire, backfeeding relay, or momentary voltage drop allowing the contactor to drop out momentarily, or anything that might interrupt power for even a moment could cause a spline or shaft problem. The interruption of power could be very small, just milliseconds could be a problem.

The check valve in the pump cannot close fast enough to prevent a fast reversal if power is momentarily interrupted.

I know this is a small pump, but we treat our controls on all pumps as follows for this very reason:

The pump starting contactor is not allowed to be connected directly to any sensor such as a pressure switch or flow switch. The contactor is locked on and held on by the PLC (or by a time delay if you do not use PLCs). If a pressure switch or flow switch or any device calls to start or stop the pump, the PLC waits a bit, perhaps around 10 seconds on stop and perhaps about 3 seconds to start. That way you get no false or fast starts and stops due to momentary blips from a sensor or start button.

PUMPDESIGNER

 

[islandeng]

PUMPDESIGNER,

There are no sensors hooked up to this system. The contactor is linked directly to the power switch which is held on, unless power is cut off, in which case it trips off and stays off.

As I noted before, the failure resembles more of a fatigue failure than an impact failure. The splines look as if they were worn off (amazing one could wear them off within a span of 3 weeks...) rather than some large impact coming and shearing them off. I suppose I have been using the wrong terminology in explaining the primary symptom with "shear" as that connotates more of a catastrophic failure.

Since we don't have any sensors, I cannot know if there had been reduced flow rate as the splines were wearing down. Please keep in mind that this is an extreemly simple system, with as few parts as possible for failure because that is how the culture works here. (It took me some time to begin to realize that mindset myself.)

I am beginning to think this is a material problem. And no one has commented about the length differences of the splines...So I take it that it is not out of ordinary.

Thank all of you for your thoughts, more are appreciated.

-loyal

 

[25362]

One could asume that splines mean crevices... and crevices in salty water invite corrosion. I suggest you bring this thread to the corrosion enginering forum.

 

[PUMPDESIGNER]

Not a metallurgist, but 3 months may be long enough for corrosion to have an effect, 3 weeks I cannot see as long enough for corrosion. I suppose that is why I never considered corrosion.

I do understand island mentality and I do respect it at times when appropriate.

If the problem appears to be wear, consider that the shaft moving up and down within the coupling would wear the splines quickly. Shock would produce an up and down motion if the thrust bearing is not tight. I have no direct knowledge about thrust bearings on those little motors, but I do know a guy who uses them in his product. He buys 100 of those little motors at a time, cuts the ends off, then machines his own thrust bearings into them. Reason is that he claims those little 4" motors have little or no ability to withstand thrust. He has been doing this for years. He has been doing it on the Franklin motors, do not know if he does it on any other brand.

We had to put in a small 4" motor last year. We were able to get an oil filled 2 hp 4" motor with roller thrust bearings. Have no idea if that thing will last or not, but we tried to get the best one we could.

PUMPDESIGNER