14 replies to this topic

[Shahine]

Dear All;

 

Please, In the vacuum distillation unit, I've a doubt to calculate the pressure drop through control valves located on Light vacuum gas oil / Heavy vacuum gas oil / vacuum residue and inlet feed to the unit.

 

Example:

 

The pump hydraulic for vacuum residue pump "see attachment", showing the pressure drop through control valve = 0.7 bar (minimum requirement)

 

But the client highlighted that the pressure drop through the control valve shall have at least 4 bar @ rated capacity.

 

Also, the existing vacuum unit, the existing control valve data sheet showing  pressure drop 4 bar @ rated capacity.

 

I don't know why ????

 

 

I want to know if there is any other process requirement that shall be considered in pressure drop calculation for control valves in above mentioned services or not????

 

 Generally I calculate the pressure drop through control valve as  0.1 X circuit pressure drop "without the valve" or 0.7 bar, whichever is greater (If upstream and downstream pressure are interdependent).

 

Or 0.05 X pressure discharge vessel or 0.7 bar whichever is greater (in other cases)

 

I can understand that the inlet feed control valve shall have higher pressure drop than above mentioned criteria to keep the pressure upstream the control valve > vapor pressure.

 

In other cases if the stream is heated in heat exchanger train as crud oil, it's better to keep higher pressure drop in the control valve to avoid vapor lock in heat exchanger train.

 

 

Attached Files

•  Sketch.pdf   33.1KB   46 downloads

[PingPong]

 Generally I calculate the pressure drop through control valve as  0.1 X circuit pressure drop "without the valve" or 0.7 bar, whichever is greater (If upstream and downstream pressure are interdependent).

 

Or 0.05 X pressure discharge vessel or 0.7 bar whichever is greater (in other cases)

I suggest you do not use those silly rules ever again.

 

To properly determine the required control valve deltaP one needs to understand the shape of the pump curve and the shape of the system curve.

 

So first of all you need to know what that 7.3 bar consists of. That will be mostly frictional pressure drop of the lines to the storage tank, plus the static height of the vacres in that storage tank.

And make sure that you know what is meant by bar. Whether that is barg (probably) or bara.

 

If I assume that that 7.3 barg is 6.0 bar frictional pressure drop plus 1.3 bar static height, then the total frictional pressure drop in your system (from vacuum column to storage) is 1.6 + 6.0 = 7.6 bar. And the static pressure difference is (0 + 1.3) - (-0.92 +0.7) = 1.5 bar.

 

Taking control valve deltaP as 50 % of friction plus 10 % of static will give:

0.5 * 7.6 + 0.1 * 1.5 = 4.0 bar

which is exactly what your smart client also prescribed.

[fallah]

Shahine,

 

What is the pump discharge pressure at rated capacity?

[PingPong]

It is my understanding that this topic is about designing a new vacuum unit, with new pumps and new control valves.

To determine the discharge pressure of a new pump requires determining what the pressure drop over its new control valve shall be to have proper control.

 

If this topic is about an existing system, then the control valve pressure drop results from the pressure balance of the whole circuit.

[Chemitofreak]

 

 Generally I calculate the pressure drop through control valve as  0.1 X circuit pressure drop "without the valve" or 0.7 bar, whichever is greater (If upstream and downstream pressure are interdependent).

 

Or 0.05 X pressure discharge vessel or 0.7 bar whichever is greater (in other cases)

I suggest you do not use those silly rules ever again.

 

Taking control valve deltaP as 50 % of friction plus 10 % of static will give:

0.5 * 7.6 + 0.1 * 1.5 = 4.0 bar

which is exactly what your smart client also prescribed.

 

 

@ PingPong

Please specify the basis of considering Control Valve del P as 50% of friction plus 10% of static.

 

@Shahine,

 

Nothing wrong is using 0.7 bar or 10% of frictional losses which ever is higher. It is general rule of thumb, that will help you size an optimal control valve. Once you have the pump curve, you can later analyse the control valve del P with respect to the pump curve

(depending upon the steepness of the curve).

 

I personally feel there is no point in increasing the head of a pump and then killing the pressure in the control valve (although you will end up in a lesser size control valve, but you are continuously wasting your energy in the pump and designing the system for a higher design pressure).

 

I feel the catch is in exchangers in the circuit (just a guess, as I do not know the whole system). The intent might be keeping the pressure in the system above the vapour pressure of the fluid in order to keep it in liquid phase, this will allow single phase in the exchanger, which is easier to design and operate than a two phase.

 

I would suggest you to revisit the system, there is something more than what you have specified.   

Edited by Chemitofreak, 02 December 2015 - 02:42 AM.

[latexman]

I kind of understand the "50% of friction" part, but not the "plus 10% of static" part, but I don't deal with large static pressure dPs in my process.  In my experience, that rule of thumb would result in too small of a Cv in a lot of cases.  My company's rule of thumb is 30% of friction, but the preferred method is to size the pump and CV at design flow at about 80% open AND at min. flow at about 20% open.  Sometimes it is impossible to meet these requirements, so engineering judgment and/or an innovative design is called for.

Edited by latexman, 02 December 2015 - 01:09 PM.

[Zauberberg]

Refer to attached article. Some great points are highlighted.

 

Attached Files

•  Avoid Pitfalls When Specifying Control Valves.pdf   520.31KB   53 downloads

[latexman]

Excellent article.  The 50% friction may not be considering the CV, and my 30% friction is considering the CV!  The devil is in the details!

[PingPong]

When designing a pumped circuit, like in this topic, the pressure drop of the control valve shall be:

 

ΔPcv = X*ΔPf + Y*ΔPs

 

Above I used X = 0.5 and Y = 0.1 which reproduces the 4 bar that Shahine's client wanted.

 

The above X and Y values were also prescribed by Shell in the days that I often worked on Shell projects, but that was in the 80's and 90's, so I am not sure what values for X and Y they specify nowadays.

 

Using a value of only 0.1 for X as some people above proposed is definitely too low.

One could argue that Shell's values for X and Y are very conservative, but in any case X should be at least 0.3 as latexman also stated before.

 

There is more science behind the above formula than most process engineers realize, but I will go into that in more detail here.

 

I merely intended to demonstrate to Shahine where that 4 bar control valve pressure drop could originate from.

4 bar may be considered rather conservative, but 0.7 bar in that system is definitely too small.

[Padmakar Katre]

Hi,

Can you please confirm the line size to storage is 10" or lower and is it really controlling circuit. I seriously doubt, the reason being in normal operation the vacuum residue will be routed to Delayed Coker and Solvent De-asphalting unit and very less flow to storage. VR being a congealing service there should always be some minimum flow/velocity kept in lines and hence the flow to storage and line size would be based on vacuum unit design turn down capacity as in the event of SDA & DCU shut down.

 

I believe the route of VR to DCU will govern the pump circuit hydraulics since this VR stream further undergoes in heat exchange with HCGO product & slip stream of DCU main fractionator. The destination pressure in case of DCU would be higher than storage tank for VR stream hence there is a logic to kill the pressure/pump head in control valve in line to storage.

 

I presume the circuit hydraulics of VDU fractionator bottom pump with destination to DCU would be be governing with control delta P 0.7 bar at rated flow and control valves in other destinations higher.

[Chemitofreak]

I agree with Padmakar, the governing circuit needs to be rechecked.

[Shahine]

Hi,

Can you please confirm the line size to storage is 10" or lower and is it really controlling circuit. I seriously doubt, the reason being in normal operation the vacuum residue will be routed to Delayed Coker and Solvent De-asphalting unit and very less flow to storage. VR being a congealing service there should always be some minimum flow/velocity kept in lines and hence the flow to storage and line size would be based on vacuum unit design turn down capacity as in the event of SDA & DCU shut down.

 

I believe the route of VR to DCU will govern the pump circuit hydraulics since this VR stream further undergoes in heat exchange with HCGO product & slip stream of DCU main fractionator. The destination pressure in case of DCU would be higher than storage tank for VR stream hence there is a logic to kill the pressure/pump head in control valve in line to storage.

 

I presume the circuit hydraulics of VDU fractionator bottom pump with destination to DCU would be be governing with control delta P 0.7 bar at rated flow and control valves in other destinations higher.

Dear Mr. Padmakar Katre

Thanks for your reply, But the VR will be delivered at unit battery limit, the battery limit pressure of VR to storage is higher than battery limit pressure to DCU or  SDA.

SO, the controlling route in my case is VR to storage and I have to size the pump to match with this higher deliverable pressure .

 

Thanks;

[Shahine]

Shahine,

 

What is the pump discharge pressure at rated capacity?

Dear Mr. Fallah;

 

I want to calculate the discharge pressure of this new pump:

 

So, if I consider the pressure drop through control valve = 0.7 bar, the estimated pump discharge pressure = 14.5 barg

And If I considered the pressure drop through the control valve = 4 bar, Estimated pump discharge pressure = 18 barg

 

Thanks;

[Shahine]

Refer to attached article. Some great points are highlighted.

Thanks for your support

[Dacs]

You have to realize that the vacuum residue has a very low vapor pressure and the exchangers are actually cooling down the vacuum residue (but not cold enough to reach pour point, but that's another topic for discussion), so there's no vapor formation on this line.

 

However, I need to highlight that you have to revisit your circuit (as others have mentioned). Unless the tank farm is located far away from your VDU, I'm having doubts that this will be your governing case. Can you afford to do full hydraulics from the pump to the end point (for all circuits) or are you just given battery limit conditions?

 

Normally, you'd want to run all cases (in your case, 3) and you'll get the highest pump head requirement among those cases. That will define your governing case.

 

To size the control valves for the other cases, you have to run your circuit considering the preliminary pump curve (which would come from the governing case). I get the feeling that the client is referring to another existing unit in which the line going to tank farm does not govern the pump sizing, and the 4.0 bar delP for that particular valve is a consequence of the pump selection.

 

In short, the 4.0 bar value may not be about CV sizing at all, but about the point I raised above.