Pumps w/VFD blocked outlet protection

[fdomin]

We have a pair of crude oil centrifugal pumps (680 m3/h - 3000 gpm) with large discharge head (192 m - 630ft )that need to be reused. They will be speed controlled by means of a VFD. Minimum flow is 1/3 of rated flow.

Spillback is not desired as return to suction vessel is impractical (vessel inlet flow cannot be modified) and power consumption will be 30% more. Also we have no flow controller to install a low flow switch to stop the pump.

Is it possible to use a high speed switch to stop the pump in the case of running at shut-off?

Speed controller will try to increase speed to achieve pressure set-point. The main problem I can forsee with this scheme is that shut-off head will also be increased leading to a worse situation.

Any advice will be appretiated.

Thanks

 

[BigInch]

Sounds like you may be a bit confused (or it could be its me that's confused).

VSD controlled pumps won't be discharging at shutoff head, assuming the VSD is not broken. VSD controlled pumps will be delivering a differential head equal to BEP head * Actual_RPM^2/BEP_RPM^2. VSD controlled pumps will increase discharge head from 0 m head at 0 RPM to 192 m at BEP_rpm.

You must set the speed to whatever you need to yield the maximum discharge pressure(=head suction head + 192 m) when you are at full flow (680 m3/h). Speed should not increase above that, since maximum pressure setting (=your speed control signal) would be exceeded, THEREFORE rpm should stay at that corresponding setting and NOT increase any higher.

At minimum flow = 1/3 full flow, the pumps will be at 1/3 BEP_rpm and DIFFERENTIAL head output by the pump will be 192/1^2*0.3333^2 = 21.3 meters or, roughly about 25 psi, 200 kPa.

If that head is not sufficient to pump the minimum flow the downstream pipeline, you have TWO choices.
1.) you MUST flow back to a tank, or
2.) you could opt to recycle back to the pumps by tying a "backspill" line from the pump's discharge header into the suction lines or suction header. If you do that, you should watch for possible temperature increases at the pump. At 1/3 speed, temperatures may get high after anywhere from 10 to 30 minutes and may require a pump shutdown. Use a high temperature indicator and switch.

Why are you worried about minimum flow anyway? If you can't pump into the downstream pipeline at 1/3 rpm, shut the pumps down! If you're starting up at 1/3 rpm, don't run at 1/3 too long and watch the temp when you do.

http://virtualpipeline.spaces.msn.com

 

[BigInch]

NOW I GET IT!

If you think you may have low flows at high rpm for some reason, do this;

Put a PT on the suction side and use that to REDUCE SPEED when SUCTION PRESSURE FALLS. At the controller, you would compare the suction PT output (say 1 at maximum suction pressure, 0 at < minimum suction pressure) with the discharge PT output (say=1 at 192 meters) and take the lower of the two outputs. Use the lowest value as the speed factor x BEP_rpm. That way, rpm will be 0 whenever you get below min suction pressure.

As for a discharge block valve closing, without a flow meter, you will have to depend on a high pressure switch override to the Discharge PT signal, to shut the pump down.

http://virtualpipeline.spaces.msn.com

 

[fdomin]

BigInch,

Thanks for your answer. I can see that I didnt explain my system correctly. I will try to explain it better.

The pump suction vessel is a preflash drum and flashed crude is then pumped to a fired heater. An existing pressure controller sets the speed of the speed controller to maintain a fixed pressure upstream of the heater inlet valves.
If the pump is blocked with its own discharge valve, pressure at the heater inlet valves would decrease and controller should react setting the speed to higher RPM´s trying to maintain the set pressure.
A Hazop study indicates that I must protect the pump from running at shutoff to avoid high temperatures an pump damage. Pump MCSF is 1/3 of rated flow and as I said, is almost impossible to install a backspill line with restriction orifice to the vessel, then I need to stop the pump in case of a blocked discharge valve.

Thanks for your advice for a PT in the suction side. We already had taught of it to stop the pump at minimum suction pressure, but not for controlling the RPM´s.

 

[BigInch]

You're welcome.

Then as I see it, all you need is a HiPT on the pump discharge going to a pump shut-down sequence, whenever the pumps start backing up the curve and start delivering higher pressure than what is needed to get the 680 m3/h into the heater under normal circumstances.

http://virtualpipeline.spaces.msn.com

 

[dcasto]

I do not see how your system will work at all. What if the preflash drum makes NO liquids, the pump will run dry and the proposed system will shutdown and the heater will fire until some system shuts it down. Next, if the prflash makes 1/4 the liquid, the proposed sytem will shutdown and the plant will stop all beacuse the system is short a few bbls/hr, thats a bigger waste than recycling a minimum flow.

I do not see a way that your system can keep pressure on the heater, its the downstream system that keeps pressure on the heater and your pump will respond to something happening PAST the heater. All heaters I know of have minimum flows through them. How will you keep a level in the preflash drum, just let the pump whatever it wants?

I believe you need to either rethink the whole thing or explain your system more.

 

[BenThayer]

if you have a scenario where you are blocking in the flow to a fired heater, i would be more concerned about a flowmeter and protection of the fired heater!

if it has the classic "outlet temperature controls firing" then no flow can cause a low outlet temp and the firing will increase and potentially cause complete failure of the heater and a major fire.

i think you may bigger fish to fry with this oil than the pump.

 

[BigInch]

Valid concerns.

dcastro, Right, and as s I always say, VSDs are NOT for every application. I also think that production would shut down well before 1/3 flow is reached, since there wouldn't be much head to get into the heater well before 1/4 flow.
I agree that including a flow control would be much better for the system, but also think that there should be an independent level and, as BenThayer notes, a temperature controller at the heater.

Now though it is a combination of outlet pressure held downstream (or at) the heater outlet plus inlet pressure that keeps the heater pressured up? It won't work with downstream pressure alone. The pump (with PT) will respond to wherever the PTs are located, not to events past the heater, except as to how those events register at the pump's PT.

http://virtualpipeline.spaces.msn.com

 

[fdomin]

BigInch/dcasto,
We are installing the preflash system between the last heat exchanger an the fired heater.
Nowadays the pressure upstream the heater valves is fixed by a PT that sets the speed of a booster pump downstream the desalter. Pressure at the heater is defined by column pressure and the valves drop pressure from PT setpoint to column pressure + heater DP.
Unit feedrate is defined with the heater pass flowrate controllers.

We try to keep existing control philosopy as much as we can, then we are redirecting the signal of this PT to the new preflash pumps, maintaining a similar configuration of existing control.
With the PF drum insertion the heater valves cannot set the unit feed directly, so we need a new feed controller upstream the PF drum. This will be a cascade control loop that sets the unit feedrate while the PF drum level modifies the flow setpoint.

A new PT will control pressure upstream the flash valve to avoid early vaporization.

The heater has its own control logic with control valves setting each pass flowrate and we do not change this configuration. If the pump stops or flow is blocked the heater will shutdown as it would do today if the feed or booster pumps stop.