Intuitive description of what causes upthrust?

Users who are viewing this thread

Reach4

Well-Known Member
Messages
38,749
Reaction score
4,400
Points
113
Location
IL
Does anybody have an intuitive description/model for what causes upthrust in a common submersible centrifugal pump? I know it happens when you pump water at an excessive rate, but I don't have a feel for why this would be the case.
 

LLigetfa

DIYer, not in the trades
Messages
7,493
Reaction score
574
Points
113
Location
NW Ontario, Canada
I know it happens when you pump water at an excessive rate, but I don't have a feel for why this would be the case.

The high rate of flow combined with low back pressure is what causes the upthrust. Many multi-stage pumps float their impellers and this allows them to "float".

I wonder if pumps that have their impellers fixed on a shaft would be less prone to upthrust?

Also wonder about systems with bleeder/snifter/check airmakers. Since the pump starts against very little head, at a very high rate of flow how is it they don't self-destruct?
 

Valveman

Cary Austin
Staff member
Messages
14,583
Reaction score
1,296
Points
113
Location
Lubbock, Texas
Website
cyclestopvalves.com
Impellers that are fixed to the shaft, benefit slightly from the weight of the shaft. Also the top end of the shaft usually hits a stop in the bushing on the discharge bowl, keeping the impellers from actually touching the top of the diffusers. The stop in the discharge is not built strong enough for the pump to run in this condition for very long. When it wears out, the top of the impellers start dragging the bottoms of the diffusers, and pump failure is not far away.

Floating impellers do not benefit from the weigh of the shaft, as they can easily slide on the shaft a small amount. The shaft can still have a lock ring that keeps the impellers from rising too high, until the same failure procedure happens as with the fixed impellers. Either way up-thrust is not good for a pump.

And yes there can be considerable up-thrust on a pump used with a bleeder system. Some wells are deep enough that the deep static water level puts enough head on the pump to prevent up-thrust. With a high static water level, up-thrust can be a problem.

Up-thrust happens on basically every pump start. Until the check valve opens, there is no head against the pump. The stop in the upper bushing is designed to handle up-thrust for short durations fairly well. It is when running a long time with too little head on the pump that up-thrust can be the most destructive.
 

LLigetfa

DIYer, not in the trades
Messages
7,493
Reaction score
574
Points
113
Location
NW Ontario, Canada
Up-thrust happens on basically every pump start. Until the check valve opens, there is no head against the pump...

This does not jive with the high velocity induced uptrust theory. There cannot be high velocity when the pump starts against a closed checkvalve since water doesn't compress. Well... at least not against a check in the pump. A check topside in the snifter is a different matter since the line is filled with air that compresses.
 

Valveman

Cary Austin
Staff member
Messages
14,583
Reaction score
1,296
Points
113
Location
Lubbock, Texas
Website
cyclestopvalves.com
This does not jive with the high velocity induced up-trust theory.

Yes it does. Water is not compressible but it can be pressurized or non-pressurized. Below the check valve, water in a submersible pump is not pressurized when the pump is off. Above the check valve, water is pressurized even when the pump is off.

At the instant of pump start, high velocity is induced so the impellers stack, and even the rotor in the motor go into up-thrust condition. When the pressure in the pump is high enough to push the check valve open, the impeller stack and rotor are pushed back down.

This all happens in a fraction of a second. But each time the pump is started the impellers, shaft, and rotor fly up, and then get slammed back down. This is another reason why multiple starts are not good for a pump and/or motor.

I am working on a case now where the installer is regularly shattering thrust bearings in a 50 HP submersible motor. It is set at 1200’ and has 5 check valves spaced evenly in the drop pipe. The manufacturers are telling him he must have a check valve every 200’. I am telling him the multiple check valves are the cause of the fractured thrust bearings.

A pump that is designed to work from 1200’ will be in an up-thrust condition until it gets 1200’ of head applied to it. Each check valve up the drop pipe isolates the pump from the 1200’ of head. Until the bottom check valve is pushed open, the pump only sees 200’ of head. Until the second check valve from the bottom is pushed open, the pump only sees 400’ of head, and so on. Not until the fifth check valve at the top is pushed open will the pump see the 1200’ of head it needs to keep from being in up-thrust.

For each check valve in the drop pipe the shaft and rotor go from complete up-thrust to being hammered back down as soon as the water hits the next check valve. Water hammer can easily fracture a thrust bearing. And the 5 check valves are causing 5 water hammer events before the water ever starts flowing out the top of the well. The thrust bearings look like they have been busted with a sledgehammer.

Having only one check valve attached directly to the pump will solve this problem. Sure the pump will go into up-thrust until this check valve is pushed open. But the short distance between the pump and the check valve will cause much less shock to the thrust bearing than check valves higher in the drop pipe. When the pump is started, each additional check valve causes the pump to go from an up-thrust condition to hammering the rotor down against the thrust bearing.
 
Last edited:

Reach4

Well-Known Member
Messages
38,749
Reaction score
4,400
Points
113
Location
IL
At the instant of pump start, high velocity is induced so the impellers stack, and even the rotor in the motor go into up-thrust condition. When the pressure in the pump is high enough to push the check valve open, the impeller stack and rotor are pushed back down.

This all happens in a fraction of a second. But each time the pump is started the impellers, shaft, and rotor fly up, and then get slammed back down. This is another reason why multiple starts are not good for a pump and/or motor.
What causes an impeller to fly up at startup? I would usually think that would be associated with pushing water downward. So I am now thinking of either
1. the higher pressure at the outer part of the "semi‐open"*impeller pushing the top shroud upward,
2. or the water moving outward hitting the lower part of the volute giving a reaction upward.
 

LLigetfa

DIYer, not in the trades
Messages
7,493
Reaction score
574
Points
113
Location
NW Ontario, Canada
The high flow rate of water can lift the impellers. These are multi-stage pumps and water from each stage is pushed up to the next stage.
 

LLigetfa

DIYer, not in the trades
Messages
7,493
Reaction score
574
Points
113
Location
NW Ontario, Canada
I am working on a case now where the installer is regularly shattering thrust bearings in a 50 HP submersible motor. It is set at 1200’ and has 5 check valves spaced evenly in the drop pipe. The manufacturers are telling him he must have a check valve every 200’. I am telling him the multiple check valves are the cause of the fractured thrust bearings...
Multi checks I can understand because you could actually develop a partial vacuum between the checks. That vacuum then causes a rush of flow until it hits the check that is holding pressure. That results in a shock wave.

At least with the bleeder/snifter/check, there is a soft cushion of air.
 

Valveman

Cary Austin
Staff member
Messages
14,583
Reaction score
1,296
Points
113
Location
Lubbock, Texas
Website
cyclestopvalves.com
The types of pumps used in these applications are “centrifugal”. They work using centrifugal force, which means they throw water outwards, not upwards. There is no downward force on the impellers. They suck in water from the inlet eye at the bottom, and throw it straight sideways into a diffuser. The diffuser works like an elbow, turning the flow of water from sideways to straight up into the eye of the next impeller. The next impeller is actually thrown upwards from the velocity of water hitting it from below.

It is the speed at the outer edge of the impeller that determines the velocity head one impeller can produce. Each additional impeller adds the same amount of velocity head in an accumulative way. Ten impellers cannot pump any more flow than one impeller. They can just build ten times more pressure or head.

Only after there is sufficient backpressure will there be downward force acting on centrifugal impellers.
 

NHmaster3015

Master Plumber
Messages
833
Reaction score
2
Points
0
Location
The granite state
Yes it does. Water is not compressible but it can be pressurized or non-pressurized. Below the check valve, water in a submersible pump is not pressurized when the pump is off. Above the check valve, water is pressurized even when the pump is off.

At the instant of pump start, high velocity is induced so the impellers stack, and even the rotor in the motor go into up-thrust condition. When the pressure in the pump is high enough to push the check valve open, the impeller stack and rotor are pushed back down.

This all happens in a fraction of a second. But each time the pump is started the impellers, shaft, and rotor fly up, and then get slammed back down. This is another reason why multiple starts are not good for a pump and/or motor.

I am working on a case now where the installer is regularly shattering thrust bearings in a 50 HP submersible motor. It is set at 1200’ and has 5 check valves spaced evenly in the drop pipe. The manufacturers are telling him he must have a check valve every 200’. I am telling him the multiple check valves are the cause of the fractured thrust bearings.

A pump that is designed to work from 1200’ will be in an up-thrust condition until it gets 1200’ of head applied to it. Each check valve up the drop pipe isolates the pump from the 1200’ of head. Until the bottom check valve is pushed open, the pump only sees 200’ of head. Until the second check valve from the bottom is pushed open, the pump only sees 400’ of head, and so on. Not until the fifth check valve at the top is pushed open will the pump see the 1200’ of head it needs to keep from being in up-thrust.

For each check valve in the drop pipe the shaft and rotor go from complete up-thrust to being hammered back down as soon as the water hits the next check valve. Water hammer can easily fracture a thrust bearing. And the 5 check valves are causing 5 water hammer events before the water ever starts flowing out the top of the well. The thrust bearings look like they have been busted with a sledgehammer.

Having only one check valve attached directly to the pump will solve this problem. Sure the pump will go into up-thrust until this check valve is pushed open. But the short distance between the pump and the check valve will cause much less shock to the thrust bearing than check valves higher in the drop pipe. When the pump is started, each additional check valve causes the pump to go from an up-thrust condition to hammering the rotor down against the thrust bearing.

Ran into a similar problem about 10 years ago. Pump at 1000', checks every 200' pulled it, removed all but the check at the pump. its still down the hole, still pumping just fine. Original installer was worried that the single check valve wouldn't be able to hold back 1000' of head pressure. I tried to explain to him that the check at the pump is holding the entire head regardless of the other checks in the line, big argument, hurt feelings......LOL
 
Top
Hey, wait a minute.

This is awkward, but...

It looks like you're using an ad blocker. We get it, but (1) terrylove.com can't live without ads, and (2) ad blockers can cause issues with videos and comments. If you'd like to support the site, please allow ads.

If any particular ad is your REASON for blocking ads, please let us know. We might be able to do something about it. Thanks.
I've Disabled AdBlock    No Thanks