Startup Bypass Loop

[timebone]

So, I've been having an incredibly difficult time getting a solid answer on this. I'm a pretty inexperienced EE so pumps are not really my specialty.

I'm a bit confused about the purpose of a start up bypass loop. These are centrifugal pumps that start against a closed discharge valve to prevent run out conditions. However there is also a bypass loop that will begin to close once the discharge valve is fully open.

Can anyone explain what this is achieving? I've never seen this before... someone has mentioned that it helps prevent deadheading, but how can deadheading be prevented if you're still starting the pump against a closed discharge? Does this bypass loop really have that big of an effect on the system curve when the valve starts opening?

I see pictures of bypass loops but they typically have this bypass loop upstream of the discharge valve. So when you start your pump, the discharge valve can be closed and correct sizing of the bypass line should put your pump in safe operating conditions.

Here is what I'm talking about. I've excluded the suction valve but there is one. You can see there is the discharge (closed when starting) and then the bypass valve (open when starting). Once the pump starts, the discharge opens. Once fully open, the bypass valve closes.

http://postimg.org/image/i05izod21/

 

[MikeHalloran]

I suspect the bypass was included to modify the system curve as seen at the pump, in order to get over a 'hump' associated with system startup.

I.e., if you block the bypass, you may blow a fuse.

Don't do that. Take a look at the pump operating points by superimposing the different system curves on the pump diagram, and noting the efficiency or power curves' locations also.



Mike Halloran
Pembroke Pines, FL, USA

 

[pennpiper]

Timebone,
I may be reading your attachment sketch all wrong, but it appears the recirculating loop is branching off the discharge after the discharge valve. If it is installed the way your sketch shows it (the bypass loop) will be nonfunctional.

Of course there is a lot of items missing from the sketch.
How about adding:
- Flow arrows, upstream, downstream and the loop
- The Suction Block Valve
- the discharge Check Valve
- The source of the suction (Tank, Pipeline, other?), Yes it can make a difference
- Line sizes

prognosis: Lead or Lag

 

[bimr]

Depending on the application, some engineers specify special surge anticipator valves. The surge anticipator valve discharges from the pump discharge header. The valve anticipates and prevents surges caused by power failure as well as acting as a standard over pressure relief valve.

See the valves in the link.

http://www.oawu.net/Pdf/CV_STM_Section_3_Applications.pdf

You need to provide further details for someone to understand your exact application.

 

[JJPellin]

We would commonly add spill-back lines like these for large, high energy or high speed pumps that cannot tolerate operation at zero flow. For example, our large charge pumps in our high pressure gas-oil hydrotreater. These are 1500 HP, motor driven barrel pumps. They put up about 1800 psi discharge pressure. If these pumps were operated against a blocked discharge, the energy input into the pump would vaporize all of the fluid and the pump would run dry in about 30 seconds. We cannot afford to risk starting this pump against a blocked valve. In addition, there is a control valve downstream of the pump that regulates the flow into the reactor. This valve could fail shut or be told to close by a failure of associated instrumentation. In addition, we cannot risk the possibility of back-flow through the pump if it were to trip and the check valve stick open. This would allow 1800 psi hydrogen to flow back into the suction drum which is not rated for that pressure or designed for that stream. To protect against this back-flow event, there are automatic, fast closing chop valves that will slam shut on a loss of forward flow. If these valves were triggered to close, I would not want the pump to run dead-headed and destroy itself. A catastrophic wreck of the pump in this scenario could result in a large release and a fire. The spill-back will open automatically if the chop valves close which protects the pump against destruction. Lastly, this process is prone to a possible run-away reaction if the ratio of oil to hydrogen is not maintained. When starting up a pump, or switching pumps, we need to make the transitions slowly and smoothly. The spill-back allows me to gradually introduce the flow from the second pump without forcing me to run the pump below the recommended minimum flow.

Johnny Pellin

 

[timebone]

Thanks for the replies guys. I'll speak with the engineers about their intended goals for this bypass line.

 

[MikeHalloran]

When conducting an industrial archaeology dig, i.e. trying to find out why something was done in a particular way, do not forget to ask the maintenance crew, who did the work, and the production crew, who saw the before and after performance.

The best sources for this sort of thing are often identified as having a 'bad attitude'. ... because they care, or cared once upon a time.

It might cost you some beer and pizza, always a good investment.


Mike Halloran
Pembroke Pines, FL, USA

 

[timebone]

Mike: I will absolutely follow up with these guys. I think it might have been a simple oversight by the automation guys.