metering pump vs. control valve in sour water stripper

[mjaeger]

We are working on replacing / upgrading one of three parallel sour water strippers in a refinery in Germany.
In the exisiting design, the sulfuric acid is dosed via a metering pump with variable stroke from an elevated tank to the sour water stripper. Sour water stripper and sulfuric acid tank are approx. 300 m apart, the metering pump is close to the sour water stripper. There is a pulsation damper on the pressure side of the pump, none on the suction side. The suction line tends to vibrate strongly. We need to dose on average 100 kg/h (55 l/h) sulfuric acid, max. approx. 380 kg/h (200 l/h). The metering pump can cover this range.
To overcome the vibration problems in the suction line we are looking at the following alternatives:
1. add a pulsation damper on the suction side
question: what size (volume) will be required for the pulsation damper?
question: will a suction side pulsation damper be sufficient to solve the vibration problems? Or should alternative 2 be preferred?
2. add a pulsation damper on the suction side AND move the sulfuric acid tank closer to the metering pump
3. add a pulsation damper on the suction side AND/OR move the sulfuric acid tank closer to the metering pump, use the metering pump at full stroke, recycle excess sulfuric acid to the tank and use a control valve to dose the sulfuric acid
question: can one control valve cover the required turn down?
question: can a DN25 control valve be used? are smaller control valves easily available and at reasonable prices?
question: are metering pumps or control valves better suited for dosing within the required range?
Aside from these specific questions we will also greatly appreciate any comments on operation experiences with sour water strippers (e.g. fouling, direct steam vs. reboiler, pH measurement, pH control, cleaning of pH measurements).
Thanks in advance for any advice.

 

[25362]

May be these threads are of some help:

thread124-144047
thread124-171887

 

[JJPellin]

We have seen similar vibration problems with pumps of this type. The resolution of the most recent problem required the addition of pulsation dampers to both the suction and discharge sides of the pump. The damper on the discharge is to smooth out the discharge flow and reduce piping pulsation. The damper on the suction side is to control piping pulsation but also to eliminate cavitation in the pump. A pump like this will cavitate even with adequate static pressure on the suction if the suction piping is too long. A simple calculation of the acceleration head based on the size and length of the suction line, the stroke rate of the pump and the volume displaced per stroke can tell if you are operating in a bad condition. You have not provided enough information to answer your question. The pump must have adequate NPSH to avoid cavitation. Moving the tank closer to the pump may help. But the height of the tank must also be adequate. The suction pulsation damper can help reduce the acceleration head, but does nothing to improve the static suction head to the pump. If the problem is related to cavitation, running at full stroke length and spilling back the excess will not help, it will make the problem worse. Cameron Hydraulic Data by C.C. Heald has an excellent section on how to calculate NPSH and acceleration head for this type of pump.

Based on the information you have provided, I would calculate the NPSH available to the pump. If acceleration head is the limiting factor, add a pulsation damper sized for the flow and pressure in your system. Any supplier of this equipment can help you choose the correct size. If static head is the problem, move the tank closer to the pump. This might solve both problems but sounds like the more expensive and complex solution. I would not consider your third option.

Johnny Pellin

 

[mjaeger]

Hi Johnny Pellin and 25362,
Thanks both for the quick feed back!
I'll check out the literature on the sour water literature thread.
I had already studied the maximize sour water feed inlets thread. It confirmed our conclusion that reboiler vs. direct steam is pretty much a tie. The required energy in form of low pressure steam is the same for both. With direct steam we lose the steam to the waste water, with a reboiler we have the excess capital investment and fouling problems in a second piece of equipment (the reboiler).
The available NPSH is 8.7 m at min liquid level in the tank (pressureless tank, min level at 3 m height, density H2SO4 of 1830 kg/m3, vapor pressure of H2SO4 practically 0).
According to a formula I found in the 27. edition from 1944 of "Die Hütte" (German engineering compendium,) the current design requires 6.3 m. The formula is
p-s = L-s x d^2 / d-S^2 x a / g
with
p-s = required suction pressure
L-s = length of suction line = 300 m
d-S = diameter of suction line = DN80
d = diameter of piston = 68 mm
a = acceleration = s / 2 x n^2
s = stroke = 127 mm (max.)
n = stroke rate = 127 min^-1
g = 9.81 m/s
max. volume displaced at max. stroke = 800 l/h
In theory the NPSH should be sufficient. In reality it is not. The pump does indeed cavitate at low liquid levels above the min. level. Which leads me to think I'm using a wrong formula or not the correct values.
Relocating the tank seems to be a worse solution than it actually is. Both the tank and the retention basin do not fullfill environmental requirements anymore and need massive refurbishing.
What I don't like about the control valves is that I think they have a lower turn down ratio than a metering pump. The rule of thumb I'm used to is that control valves operate reliantly from 20 to 110%, i.e. in this case 40 to 220 l/h, which would be sufficient. If we use control valves, I would favor a pressurized tank instead of creating pressure via pumps (metering or any other). Of course a pressurized tank for acid has its setbacks too. However, a 50 m3 tank is large enough.