Startup Bypass Loop - Follow up

[timebone]

Hey guys,

So I just wanted to post a follow up to the thread I made a few days ago.

So it turns out, the design was incorrect: the pump discharge valves were supposed to open and then you would start the pump with the bypass loop open. So thanks for the help with that.

My follow up for this is as follows: how does this startup bypass loop prevent any sort of run out conditions during pump start up? Lets say that the line downstream is not fully pressured up. When I start that pump, won't I still face run out conditions regardless of if I have a bypass loop present? It seems like although this bypass loop can be used to avoid deadheading, I don't see how it would affect run out conditions. I feel like I might be missing something...

Thanks guys.

 

[MikeHalloran]

You didn't have to start a new discussion for that; consider putting the above in a reply to the original discussion, then Red Flagging this discussion and asking that it be removed.


Mike Halloran
Pembroke Pines, FL, USA

 

[MortenA]

Isnt there a control valve or an orifice in the start-up loop? If there is then you wont have run-out conditions. But you will see an increase in fluid temperature and therefore you should use the loop for too log before starting to flow in the main line. Once youhave some flow it will (most likely) not be a problem.

 

[LittleInch]

On the basis of what you have provided, you are not missing anything and for a liquid pump, your "start up loop" would not prevent run out of the pump on start up if your discharge conditions don't initially provide the back pressure that it would in normal operation.

However without seeing the pump curve, actual start up condition, any control valves downstream, it is difficult / impossible to comment further.

Start up is often an overlooked operation, but so long as the condition is quite short, in reality pumps survive this quite well. Occasionally the pump will trip on high amps and if the pump is showing signs of distress, (vibration, high amps, damaging seals etc) then you probably need a flow restriction (control valve) downstream which limits flow during the start up then opens fully.

This is often why liquid pumps start against a closed valve which somewhat crudely undertakes this operation (isolation valves are not good control valves), but can reduce or smooth out the initial high flow surge into an unpressurised system.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[timebone]

Mike: Sorry, won't happen again.

MortenA: Yes, there is a restricting orifice in the start-up loop. But again, isn't this line irrelevant if the downstream line doesn't have a lot of back pressure? Even if that bypass loop is fully closed/fully open, wont I still see run out conditions if I start the pump with low back pressure downstream?

LittleInch: The only control is the pump speed via VFD control. This VFD is controlled off of discharge pressure, and the pump gets a starting set point of basically 100% of the rpm range. But discharge pressure isn't pump head... so I don't know how controlling discharge pressure would really affect run out either. But even so, if there is no back pressure - regardless of the speed of the pump - the line impedance downstream of the pump is still going to be low... It seems like this bypass loop avoids deadheading but then how are you avoiding run out while starting? I'm wondering if they start at max speed to fill up the line downstream faster and reduce the run out conditions faster.

How would the pump curve aid you?

 

[JJPellin]

The spill-back helps to avoid running the pump at stonewall (run-out, end of curve) by the proper use of procedures. The pump is started up with the discharge blocked in and the spill-back open. The discharge valve is opened gradually. The spill-back is closed gradually to maintain the total flow through the pump within an acceptable range. As the downstream process is flooded and pressurized, the discharge valve is opened more and more and the spill-back is closed. By using the spill-back and the procedure, the pump can be operated within the recommended flow range and the downstream process can be flooded and pressurized without ever running the pump at excessively high flow.

Johnny Pellin

 

[LittleInch]

With respect, I disagree with JJPellin. A "spill back" or recirculation line helps to avoid the pump running at low flows or no flow due to valve closure or rise in downstream pressure higher than the pump discharge pressure. It does not normally do anything with regard to end of curve or run out. If the downstream pressure is low and hence flow through the pump exceeds its minimum flow, the bypass line should then close and hence flow through the pump is dependant on the downstream pressure. If you look at the other post from the OP, he has quite as strange piping layout where the re-circ line is downstream [sic] of the pump isolation valve.

timebone. You say that the pump VFD gets a start point of 100%. Do you know this or is this a guess? There may be either a specific start-up sequence where pump speed is slowly ramped up until it reaches the required set point or there could be a feedback loop whereby the speed is limited by the amps going into the motor or maybe a flow input as an overide. I believe you are correct and that the bypass loop does absolutely nothing for you in protecting the pump on start-up from high flow and if the bypass line starts open (you don't say what type it is or how it is controlled) then would make things worse not better (compressors are different and they need to start with a bypass open, but we're talking liquid pumps here).

How do you avoid "run out" when starting? - Control the speed of the motor to limit flow / power / amps to the max of the pump. Control the flow with a control valve to do the same thing. If you slow down opening of the discharge valve (you don't say if it is actuated or not)- in fact you've not given any information on size, flow, fluid, power, motor size - then this could have a similar effect.

It all depends on how big your downstream system is, how long it takes to stabilise flow, what other control functions are in the VFD, how big your system is, how much spare the motor has for running at end of curve conditions etc etc.

oh by the way, discharge pressure is pump head - just expressed as pressure instead of metres or feet

The pump curve would help you to se eif in fact your pump was going off the end of the curve or not. For that you would need to record pressure as a minimum and flow if possible during the start-up mode. It would also tell us what sort of size we're talking about here. A 25kW motor / pump is less likely to have serious issues compared to a 500kW unit.

Bottom line is that from the low amount of info you've provided, your re-circ line is in a strange place and you don't seem to have any control over flow in the event that on start-up or during operation your downstream line has low pressure compared to "normal" operation.

How about you give a better description of what this operation is, how it works, what the pressures and sizes are, what your current issues / problems are.... Then we might get somewhere

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[timebone]

LittleInch:

I'm not trying to be intentionally vague here. I think you've generally gone above and beyond answering my question. I really just wanted to confirm that this bypass loop has no effect on the start up conditions.

But you know what, now that I think of it - you're right. If you keep the VFD speed low, you will be in a sense "limiting the amps to the motor". However, I think what I meant by my statement was that the pump is still "running out" in a sense. I say this because regardless of whether the pump is max speed or min speed, the pump will still be in a position where there is little back pressure and you'll still be at the end of your pump curve. But yes, you're right. I guess if the pump is sized in such a way that running it at minimum speed at the end of the pump curve doesn't cause concerns, then I guess it's fine.

I'm not guessing that the pump should be started at max speed - it's mentioned in the control narrative. Like I said, I am assuming maybe this limits the time you spend being at the end of your pump curve simply because the line will flood much faster at max speed than min speed. But like you pointed out, the motor amps will be much greater with max speed than minimum speed.

I always thought pump curves were based off of differential pressure/differential head. I'm talking about just the discharge pressure here.. I understand that I could say discharge head, but those aren't how the pump curve is generated, are they? What I was trying to say is that if we control of discharge pressure, that doesn't directly control where we are in our Pump curve. Sicne we also need to consider the suction pressure as well.

 

[JJPellin]

The spill-back loop cannot automatically protect the pump against operation at excessively high flow. But, the installation of the spill-back allows for the use of a procedure where the operators can start-up the pump manually in such a way as to prevent operation at excessively high flow. Without the spill-back loop, it would very difficult to get a high energy pump up and running while flooding and pressurizing the downstream system. But most of my pumps are single speed, so I am not as familiar with the value of a variable speed in this situation. LittleInch is likely much more experienced in this situation because of his pipeline background. In a refinery, it is rare to have a need to operate in this situation. We almost always have a control valve downstream of the pump that can limit the flow and control to a set-point.

Johnny Pellin

 

[LittleInch]

timebone,

Yes pump curves are based on differential head, so yes if your inlet pressure varies then at the same speed, the outlet head / pressure will also vary.

As I said earlier, it all depends on how long the pump takes to pressure the downstream system. For a short plant type piping system, probably not long enough to cause any serious damage and normally the start-up system allows for short periods of high current. For something like a pipeline though it could take a long time and lead to problems.

JJPellin,
I agree with you that in some cases you need to install a bypass loop, however the pre-requisite is a d/s control valve able to ramp up flow in a controlled manner or limit max flow for an extended period if the d/s pressure is not there for whatever reason - the OP doesn't seem to have one of these.... I do like the look of the ARV valves I've seen recently, especially the self contained ones. My normal position is that someone needs to justify putting in a recirculation line, not just install them as a default without understanding why they are there. Some high pressure, large pumps, especially with DOL starters probably do need something as running at no / low flow for even a short period can boil your liquid - I just try and only put them in when needed.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Pumpsonly]

timebone,from the diagram you posted in your original posting, your spill back or by pass loop is in the wrong position. It should be connected between the pump out let and the discharge valve. If your discharge line is empty or without back pressure, you start the pump with closed discharge and opened by pass loop. When the pump starts ,the by pass loop will provide the required minimum flow specified by the pump vendor. The by pass loop must be correctly sized to provide the necessary resistance. When the pump pressure built up, then your open the discharge valve slowly to fill up the pipe line to prevent run out.