Gravity flow line backing up 1

[Rig_man_31]

Hi pipe gurus,

Hydraulics is not my area of work but I have a problem my company has been experiencing.

We deposit multiple slurry streams into a drop box where it then travels by gravity for 4 miles.

We find that the drop box is overflowing quite often.

Along the line there are three pressure transmitters. At mile 0.9, 1.8 and 3.
The pressure transmitters at 0.9 and 1.8 read a vacuum of about -8 psi. Mile 3 reads positive pressure, but nowhere near pipeline pressure rating. We will observe backing up of the drop box followed by a spike a few minutes later in pressure at Mile 3. The drop box level will decrease and then Mile 3 will be decrease also with a few minute lag time.

If downstream transmitters are under vacuum is that an indication it is under slack flow conditions in those areas and the bottleneck is upstream?

In between the inlet and Mile 0.9, is there is a bottleneck? The inlet doesn't have fantastic grade. About -1.5% to 2% for the first 0.2 miles.

The drop box discharge line is buried deep and there isn't really an option to upsize the line. However, we could direct some of the slurry streams further down the line (line grade is about 6% here) and reconnect where the pressure transmitters are reading a vacuum. As my understanding was that in slack flow conditions the pipe is not full and their should be excess capacity there.

 

[Snickster]

What is the line size?
What is the flow rate.
What is the elevation difference between the discharge box location at the locations of the 3 pressure transmitters and at the end of the line.
Does the end of the line discharge to atmospheric pressure or if not what is the discharge pressure at the end of the line. If discharge is to a tank what is the elevation of the liquid level in the tank.
What is the slurry? Do you have any specific properties of the slurries?

 

[LittleInch]

Temperature of the slurry?

Is the pipe in the drop box always submerged or can you get air in during operation?

Do you have a time vs pressure graph?
Does the pressure fluctuate?

That low slope at the start might just create slow flow and slurry doesn't like slow flow. So the flow slows down, some air or gas appears and until the negative pressure increases to the point where the flow suddenly speeds up again the drop box backs up. Maybe.

Let me think about this a bit more but if you can answer the questions would be good.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[katmar]

When I have come across this type of surging flow it has been caused by vapor locks occurring at local high points. The fact that you have vacuum conditions in the upstream part of the line indicates that there is probably air entrainment at the drop box and if there is a high point anywhere after the mile 3 pressure transducer the air could accumulate there.

The best published description of this that I have seen was in the paper "Designing piping for gravity flow" by PD Hills in Chemical Engineering Sept 5, 1983. This paper is available for download from the Engineering.com web page. See figure 2 where this cyclical behavior is illustrated well.

Since you cannot access the discharge line from the drop box to make it larger and self venting your best bet may be to install automatic air vents at any high points after mile 3. You are sure to be warned against these vents by someone who has seen them leak and an alternative would be to install sufficiently tall vents. Once you have installed the vents the pressures that will develop will be much less so it is difficult to say how tall they need to be. Remember that a vent like this works like an air lift pump so the diameter has to be large enough for it to not function like a fountain.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[LittleInch]

That's a great paper.

it's not easy from afar to see what is happening with no data, but it does sound essentially like the pipe is too small top act as a open channel flow or maybe it acts open channel for some periods of time when slurry flow is low? Is the slurry flow variable?

Then the transition from open channel to fully flooded results in a number of "hiccups" along the way which causes the back up of material until you manage to achieve a steady state flow in a fully flooded condition.

A pressure of -8psi shouldn't be enough to draw a vacuum unless the slurry has volatile chemicals or is very hot, but could be drawing air out of the slurry.

The negative pressure will be helping to increase flow in the initial section.

To help further I think we need a detailed plan and section of this pipe to understand where slopes increase, where are the high points, what is the liquid level of the slurry in the drop box.

A second drop box further down the line may help, but we all need to see the profile, sizes, flows etc.

So what happens when it over flows? Do you reduce slurry flow until it goes down then start again?
Does it "settle down" after a while or ??


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Yes your flow rate is limited by the marginal slope at the beginning of the pipeline. Distributing flows through more branches downstream probably will not help. If you need more flow capacity, but cannot increase diameter or slope, two alternatives remain. Consider adding a looped line in the low sloped area, then join the two pipelines where slope increases. Otherwise, you need more initial head.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Rig_man_31]

Looped line is essentially what we would be looking at doing. There are two streams that deposit into this drop box. We are looking at no longer having one of them deposit into the drop box and instead extending the line downstream to 0.9 mile and then having it tied back into the main line where the pressure gauges are reading a vacuum. After the initial 0.2 mile section of -1.5% slope, the slope increases to -5.5%.

 

[LittleInch]

One thing I forgot to ask - is the "slurry" all the same?

Slurry movement by pipeline is a strange thing as its a non Newtonian fluid and small changes in composition can make big impacts in the hydraulic losses / viscosity.

So if the drop box mixed slurry changes a lot, this can be a big change to the capacity of the slurry line.

I would have thought for a relatively low slope pipe you would want to run this as open channel flow which is basically up to about 60% of the pipe area.

Once you get higher than that you run into slugs, surges and bottle necks which can result in some very strange issues.

It might sound odd, but maybe a flow control on the outlet would actually get you more capacity using something line a PE squeeze valve.

But the profile is really needed here.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[LittleInch]

The point about that though is that the extra flow will reduce or perhaps eliminate that negative pressure and then the flow capacity / flow rate in your initial system would reduce maybe to below the minimum velocity you need in slurry lines to prevent settle out?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

As requested above, what are
flow rate?
Pipe diameter?
when you see those pressures.


You can try the gravity flow pipeline calculator

https://www.omnicalculator.com/physics/pipe-flow

You will need an "effective viscosity" for your slurry.

Try running it for each section of pipeline lengths and slopes. The capacity of the pipeline will be the lowest number you get.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[hacksaw]

Sounds like you are dealing with solids build up, that once large enough to block the pipe, the gravity head increases and flushes it out, with normal pressures soon afterwards.

 

[LittleInch]

Could be hacksaw - we just don't have enough data. Could easily be a slow build up of solids at that sort of slope angle then the DP increases unit it finally moves and a slug of solid ish material passes down the line. Or just an air pocket not moving as the pipe entrance becomes submerged?

Who knows?

That initial low slope for 3-400m though is almost certainly the key. That's just not steep enough. IMHO.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Snickster]

My guess based on what you described about the system, is that prior to overflow the flow reaches a very minimum value such that you loose liquid level in the vertical downcomer attached to the bottom of the drop box, air enters the system, and vacuum is broken downstream of 0.2 mile mark as measured by pressure transmitters at miles 0.8 and 1.8. Note that up to the 0.2 mile mark the pipe is sloped up and there is a high point at the 0.2 mile mark before the pipe slopes down at 6% slope. Therefore under normal steady state flow there is a vacuum developed downstream of the 0.2 mile mark.

The loss in vacuum reduces the overall flowing capacity of the pipe so as soon as flow commences at normal rate or greater the pipe cannot flow what is entering the drop box until a time has passed and the vacuum reforms. I assume that the air trapped in the system is eventually pushed out since the flowing velocity and the available overall head of about 1000 feet from inlet to outlet is like a pumped system where the air can be forced out.

Once the vacuum reforms the level in the drop box drops and steady state steady flow conditions will then commence.

If my assumptions are correct I believe the fix would be to include level control of the drop box so it cannot be run dry and air cannot get into the system. This would require putting a control valve at the end of the line controlled by level in the drop box. Perhaps there is a way to transfer the level signal to the control valve/valve controller wirelessly.

 

[1503-44]

Not close to vapor pressure, so no volume separation, with overflow at the box. I'd guess the flow rate in the 5% region is too fast, evacuating the steeply sloped pipe at least partially and air is entering from the discharge end, but there is a full flow in the flat area blocking any further ingress of air into the upstream flatter slopes. So the air stops there at the change in slope and does not affect anything upstream. The flat slope is the "bottleneck". Again, indicating an increase in diameter, or loop that flat segment of pipe.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Snickster]

If the piping were flowing partially full downstream and air were getting in from the discharge end then I would think the pressure gauges would be reading 0 psig approx. Seems like the piping is flowing full under normal steady state flow. Anyway it is just a guessing game without all the details.

 

[LittleInch]

Rig man - We are all waiting for more information to help you.

However I think most of us are honing in on this initial rather low slope section.

I think you have a few options but don't know how feasible these are.

1) Build the walls of the drop box a lot higher by maybe 3 or 4 m?
Is the slurry pumped in?
2) Make a new drop box a lot closer to the start of the 6% slope at the 0.2 mile mark

I don't think looping the line will work as you risk lowering the velocity in the first line and getting drop out / blockages

Slurry flow needs to be kept moving at a decent velocity, usually >1m/sec and for gravity systems you need to decide are you going open channel or full pipe. Systems which can vary between the two are problematic when the changeover occurs from one to the other.

I assume / hope the line slopes continuously down as high points are a big issue, but until we get your profile no one can help you.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[katmar]

I was probably the lone voice pointing a finger at the downstream end of the line but if it is an air accumulation problem it could very well be occurring in the first section of the pipe. I missed the fact that the pressure decrease at mile 3 occurs *after* the decrease in level in the drop box. My experience is that surging flow caused by air accumulation results in very regular cycling, so a 3 minute lag of the pressure at mile 3 could equally be a 3 minute lead if the cycle time is a regular 6 minutes.

One factor that points to an air problem occurring near the start of the line is that with a pressure of -8 psi the volume of air would be double the atmospheric volume whereas towards the end of the line the air would be compressed and occupy a much smaller volume.

So I agree with LittleInch that we have got as close as possible to a solution with the available information and we need a bunch more data if we are to progress further. The questions I would have would be

1. Does this problem also occur at low flow rates or is there a threshold flow that must be exceeded to initiate this behavior. Can one of the streams into the drop box be temporarily stopped to test this. I would not build a loop system without first testing the behavior with a lower flow in the first section.
2. How regular is the cycling
3. What is the normal flow rate
4. What is the pipe size
5. Is there any back pressure at the discharge point
6. What is the particle size and density of the slurry
7. What is the profile of the line - are there high and low points. Are there any drains or vents at these points that can be used to check for air or solids accumulation.
8. Do the pressures at the first two points remain constant when the pressure at point 3 fluctuates.


Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[1503-44]

I think the fix options remain the same, no.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."