22 replies to this topic

[go-fish]

Hi,

I am new to process systems design. I have a couple of fundamental questions:

1) When does a pipeline experience a pump shut-off condition?

2) I have 2 X 100% centrifugal pumps at the outlet of an atmospheric storage tank which pumps the fluid to another atm tank located at 20 m higher elevation. Pump differential is 4 bar and max. suction pressure is 1 barg. There is a block valve at the inlet of the second tank. What is the design pressure in different sections of this system?

Thanks

[S.AHMAD]

Hi

1. Pump shut-off normally considered when any downstream valves are closed. The valves can be block, check or control valves.

2. The spec break normally downstream of the last valve that is nearest to the downstream tank. However, for low system pressure we simplified by using one design pressure over the piping system. In your case use flange rating of 150# over the pipe length and determine the pipe thickness based on the corresponding maximum allowable pressure of the flange which is a function of temperature.

[go-fish]

Thanks for your reply.

In the P&IDs for multiple pumps in parallel, I also see spec break(if neccesary) or design code break upstream of pump suction block valve. Why do we consider this scenario?

Also, if my destination is significantly higher or lower in elevation, do I need to subtract or add static head to pump shut-off pressure to calculate the design pressure at the section?

[fallah]

1-Spec break upstream of pump suction block valve is due to considering possible back pressure of working pump on stand-by one while suction valve of stand-by pump is closed.

2-Design pressure at which section? Anyway the static head should be considered in calculation of design pressure in pump shut-off case.

[go-fish]

1) How do we have a backpressure situation when we have a check valve in the pump discharge? Lets say there are 2 X 100% pumps and a last gate valve somewhere in the common discharge line gets blocked. Does a check valve stops only back flow but transfers the back pressure?

2) I meant section of piping between last gate valve (near to tank) and tank nozzle. Anyways, you answered my question. Thanks.

3) Since, shut-off pressure is experienced only during no flow conditions, can't the FIC in the discharge send a switch-off signal when there is very less forward flow(thus indicating there is an inevitable blocking in system). This way, the pump can be switched off before actually experiencing no flow conditions and shut-off pressure. Can this make a case for designing the system for lesser design pressure rather than pump shut-off? I understand in process units there are short pipe runs but in offsites, there are usually long pipe runs. Does doing something like this justify piping cost savings or it is a safety issue?

[S.AHMAD]

1. Spec break upstream of valve - This should not happen. I believe the designer was not aware of the danger and this normally happened in old design. I have personally encountered many of these situations during HAZOPS session. We normally recommend for the valve to be CSO or replace the valve rating to be in line with the pump shut-off pressure.

2. For the section upstream of the out most (last) valve, just use the pump shut-off pressure. However, we need to consider the maximum possible suction pressure(e.g. if the suction is from an atmospheric tank use maximum possible height of the suction tank, if the suction is from a pressure vessel with PSV we need to consider the worst case of psv pop due to liquid overfill). For the section downstream of the last valve, the design pressure shall be the total of elevation, frictional loss at maximum flow plus safety margin (e.g. 10% or 25 psig whichever is greater). Pump shut off pressure is not included for this section.

[S.AHMAD]

Piping design pressure lower than pump shut-off - These situations normally occur when we are using an existing pump for new service. Relying on FIC for overpressure protection is not good enough (unreliable). The goof practice is to install PSV (upstream of the first discharge valve) with set pressure the same as the piping design pressure. The PSV discharge can be routed to suction tank if this is not possible, the discharge can be routed to the suction line nearest to the tank (preferably upstream of the first valve).

[go-fish]

My last block valve is tank maintenance valve at the inlet nozzle. If there is an elevation difference of 100 m between pump and the atmospheric tank, can this valve have a design pressure of pump shut-off head minus 100 m during the blocked case?

[S.AHMAD]

We normally call that valve an "isolating valve". The design pressure rating is just the pump shut-off pressure regardless of downstream tank elevation. Since the maximum possible pressure is the pump shut-off pressure. In other words under the situation that isolating valve is closed, the maximum possible pressure that valve and upstream piping is exposed to is the pump shut off pressure. The pressure just downstream of that closed isolating valve is the tank elevation. The design pressure of that valve & upstream piping is not the differential pressure between upstream and downstream but the differential pressure between the line and the surrounding.

If you are interested to determine the minimum thickness of the valve disc then we can use the differential pressure between upstream and downstream.

During normal operation, regardless of tank elevation, the pressure any where along the piping is always lower than the pump shut-off pressure. so the valve and the upstream piping must be designed based on pump shut-off pressure NOT the difference between upstream (pump shut off) and downstream pressure.

Hope the above clarifies your concerns.

[fallah]

Yes that valve can have design pressure you mentioned.But if there isn't economical concern (due to size,material,...of the valve) it is better to take its design pressure equal to discharge line.

[fallah]

1) How do we have a backpressure situation when we have a check valve in the pump discharge?

Check valve leakage/malfunction/failure!

[go-fish]

There is an economic reason in my case. It can bring down a 24" gate valve from 300# to 150# pressure rating.

[fallah]

There is an economic reason in my case. It can bring down a 24" gate valve from 300# to 150# pressure rating.

I don't know about :

-The service
-The standard of mentioned valve
-Cost difference between two ratings
-If 24" valve 150# meets the design pressure (locally in 100 m above the pump)/design temperature of the line
-If 24" valve 150# is available (allowed to use) in relevant project PMS (PIPING MATERIAL SPECIFICATION)

Based on above and considering huge elevation difference between pump and valve,you can make a decision to select proper valve rating.

[go-fish]

What is the design pressure of suction block valve when there is a single pump? Pump shut-off or Source Vessel design pressure?

[demank]

For Design Pressure of suction block valve,
As long as the source tank is atmospheric, it is recommended to use #150 rating which have design pressure around 250 psig (15 barg).

[fallah]

Anyway,it is a good engineering practice to take the class of suction line same as discharge line.

[demank]

Thanks for your tip, fallah. Does this mean that it is preferable to have the same rating for the pump discharge as with the pump suction?

I would be be gratefull if you share with us the reference for this, because this is new to me. (In my understanding before, if pump suction line rating is #150, the pump discharge line rating may be #300)

[fallah]

(In my understanding before, if pump suction line rating is #150, the pump discharge line rating may be #300)

In normal operation you are right,but the suction line may subject to discharge line pressure due to surge happening,stopping one pump of two parallel operating pumps,......

Therefore usually both suction/discharge lines have the same class (rating,thichness) even when the pump takes suction from atmospheric tank.

[kkala]

Anyway,it is a good engineering practice to take the class of suction line same as discharge line.....
...the suction line may subject to discharge line pressure due to surge happening,stopping one pump of two parallel operating pumps,..Therefore usually both suction/discharge lines have the same class (rating,thichness) even when the pump takes suction from atmospheric tank.

In 2006 I read among other Client standards, without explanation: "In general design pressure of suction piping shall be same as that of pump discharge piping". My understanding justifies this statement in some cases only, as follows, referring to the sketches of the attached pumps.doc.
1. In case of a single pump, shutoff pressure cannot "approach" suction line. Suppose that a PSV from discharge is connected to the suction, and valve B gets closed. Apart from overheating, suction pressure remains low if A is open. If A gets closed, pump will cavitate and assumingly no head is developed (see sketch).
2. In case of two pumps of common suction and discharge (operating / standby), as shown in the relevant sketch of pumps.doc:
2a. Closing valve B at immediate discharge of operating pump will not transfer shutoff pressure to the suction.
2b. Closing valve D can pass shutoff pressure to the suction of idle pump up to the closed valve C1, but not between C1 and A.
2b1. Suppose B1 and C1 not closed (as they should have been), suction pressure remains low (valve D closed) in case that valve A is either open or closed (risk of overheating due to recirculation).
Water hammer affects both suction and discharge by an additional oscillating pressure and can be practically same at discharge or suction near the pump (e.g. on sudden closure of B or D). So it may be disregarded, considering it as a constant pressure in addition to shutoff pressure.
So the statement does not seem to apply in cases 1, 2a, 2b1; in case 2b it does not seem to apply for the suction line up to pump house, which can be long for petroleum tanks being in dikes.
Is there something wrong in the above? Any advise / explanation would be welcomed.

Attached Files

•  pumps.doc   25KB   198 downloads

Edited by kkala, 21 April 2010 - 03:42 PM.

[demank]

In 2006 I read among other Client standards, without explanation: "In general design pressure of suction piping shall be same as that of pump discharge piping".

Thanks for the broad explanation,
Then the conclusion is "IN GENERAL".
For above case, when discharge pump required #300 rating,
If single pump running, the suction pressure may be #150, rite?
Otherwise, if pararel pump installed, the suction should be #300 but common suction line still #150.

[fallah]

All above explanations are applied for the conditions you specified such as :existing PSV in discharge line that is connected to the suction (now is strongly asvised to avoid),recirculation line to be provided,...., that may wouldn't be applicable in the case of one of such conditions to be failed.

Therefore, it is better for the suction line (at least from upstream of suction valve) to have the same design pressure as discharge line.

[go-fish]

Therefore,it is better the suction line (at least from upstream of suction valve) having the same design pressure as discharge line.

I think the above is valid when there are multiple pumps operating in parallel. For a single pump case as shown in first sketch of pumps.doc, the suction line design pressure can be same as the design pressure of the upstream equipment.

[kkala]

I would specify same design pressure in suction and discharge piping, unless serious savings could be obtained (not usual). In the latter case I would justify the choice, making sure that the specific part of suction line can never "feel" shut off pressure (or some surge pressure exceeding the strength of piping class).

This because the rule of "equal design pressure for suction & discharge" tends to be a widely accepted practice. Accidents can happen even when all safety rules are respected. But if you have followed all codes and practices, you may be deemed "not faulty". This would be more difficult in case that a widely accepted practice has not been kept, even if not justified.