OilToil
Chemical
- Oct 21, 2009
- 15
Basic P&ID attached
I work in a small gas plant, and a while back one of our two demethanizer bottoms pumps went out (only 1 normally runs at a time), and we have been running on just one ever since. The two pumps were multi-stage vertical can pumps from 1948, and we couldn't find anybody that could repair/replace parts for the one that broke, so we specced out a centrifugal to replace it. We got the new pump in, installed, ready to go, and at that time everything was still working fine. The old pump was switched off (but not blocked in) and the new centrifugal pump was switched on, built pressure for a few moments, and then the pressure dropped off. Ever since then, neither the original can pump nor the new centrifugal have worked well - the pump will operate for a short time, and then vapor lock. When the vapor is vented, the pump will pump again for a little while.
I have been trying to troubleshoot this, but nothing I've done has resulted in stable operation. I'm thinking the most likely thing at this point is when the new pump was switched on we dislodged something in the tower bottoms and have a blockage in the suction line, or we are unknowingly deadheading the pumps.
Details of what's going on:
The tower bottoms liquid level has a level controller that operates a control valve on the pump discharge. When the level gets low, the discharge valve closes. Upstream of the control valve is a 1" min flow bypass line back to the tower. This line is always open, and prevents deadheading of the pumps.
The Demethanizer runs at 180 psi, and the pump discharges to a system at ~320 psi. The pump curve for the can pump runs at a nice spot on the curve, making about 75 gpm. Since our production rate is only ~25 gpm, the level control valve normally runs mostly closed, and the bypass line flows the excess flow back to the tower.
The new centrifugal pump runs at a flat spot on the curve, and the operating point has only a few pounds to deadhead (I know that is not ideal, but this pump was the only one we could get in a short time - better solution is on order). The new centrifugal should run about 60 gpm putting up 130 psi differential.
Starting from a high tower bottoms level, when the control valve is in automatic, the pump will pump down until the set point is reached. At this point, the level control valve pinches shut, and a short while later, the pump will vapor lock.
If the level control valve is shut in and bypassed, the pumps will pump until they vapor lock when the tower bottoms level drops below about 10 ft of head (~=NPSHr)
From this it would seem that there is only a problem when the outlet LCV is shut (insufficient min flow bypass) or when the the NPSHa is too low. However, the min flow bypass line has a gate valve in it, and no matter what position we put it in, the same thing happens. Keep in mind, nothing on the remaining can pump was changed, and it operated perfectly until the centrifugal pump was briefly run.
The tower bottoms temperature is 60-70F, but the liquid outlet line on the tower occasionally ices up on the outside. I think running the new pump, perhaps flowing a higher rate than the old can pumps could run, may have dislodged something into the tower bottoms outlet line, restricting the suction and flashing the bottoms liquid as it flows out.
The obvious cause of the problem would be unrelated to running the new pump, but the obvious solution doesn't fix the problem, and the problem only appeared the first time the new pump was run.
Anyone else have any other ideas?
Has anyone ever backflowed into a tower liquids outlet to try to remove an obstruction? If I did this with something that wouldn't be a problem in our process, at a rate that wouldn't lift packing/chimney trays/other internals, could that help?
I work in a small gas plant, and a while back one of our two demethanizer bottoms pumps went out (only 1 normally runs at a time), and we have been running on just one ever since. The two pumps were multi-stage vertical can pumps from 1948, and we couldn't find anybody that could repair/replace parts for the one that broke, so we specced out a centrifugal to replace it. We got the new pump in, installed, ready to go, and at that time everything was still working fine. The old pump was switched off (but not blocked in) and the new centrifugal pump was switched on, built pressure for a few moments, and then the pressure dropped off. Ever since then, neither the original can pump nor the new centrifugal have worked well - the pump will operate for a short time, and then vapor lock. When the vapor is vented, the pump will pump again for a little while.
I have been trying to troubleshoot this, but nothing I've done has resulted in stable operation. I'm thinking the most likely thing at this point is when the new pump was switched on we dislodged something in the tower bottoms and have a blockage in the suction line, or we are unknowingly deadheading the pumps.
Details of what's going on:
The tower bottoms liquid level has a level controller that operates a control valve on the pump discharge. When the level gets low, the discharge valve closes. Upstream of the control valve is a 1" min flow bypass line back to the tower. This line is always open, and prevents deadheading of the pumps.
The Demethanizer runs at 180 psi, and the pump discharges to a system at ~320 psi. The pump curve for the can pump runs at a nice spot on the curve, making about 75 gpm. Since our production rate is only ~25 gpm, the level control valve normally runs mostly closed, and the bypass line flows the excess flow back to the tower.
The new centrifugal pump runs at a flat spot on the curve, and the operating point has only a few pounds to deadhead (I know that is not ideal, but this pump was the only one we could get in a short time - better solution is on order). The new centrifugal should run about 60 gpm putting up 130 psi differential.
Starting from a high tower bottoms level, when the control valve is in automatic, the pump will pump down until the set point is reached. At this point, the level control valve pinches shut, and a short while later, the pump will vapor lock.
If the level control valve is shut in and bypassed, the pumps will pump until they vapor lock when the tower bottoms level drops below about 10 ft of head (~=NPSHr)
From this it would seem that there is only a problem when the outlet LCV is shut (insufficient min flow bypass) or when the the NPSHa is too low. However, the min flow bypass line has a gate valve in it, and no matter what position we put it in, the same thing happens. Keep in mind, nothing on the remaining can pump was changed, and it operated perfectly until the centrifugal pump was briefly run.
The tower bottoms temperature is 60-70F, but the liquid outlet line on the tower occasionally ices up on the outside. I think running the new pump, perhaps flowing a higher rate than the old can pumps could run, may have dislodged something into the tower bottoms outlet line, restricting the suction and flashing the bottoms liquid as it flows out.
The obvious cause of the problem would be unrelated to running the new pump, but the obvious solution doesn't fix the problem, and the problem only appeared the first time the new pump was run.
Anyone else have any other ideas?
Has anyone ever backflowed into a tower liquids outlet to try to remove an obstruction? If I did this with something that wouldn't be a problem in our process, at a rate that wouldn't lift packing/chimney trays/other internals, could that help?
