line and pump hydraulics 4

[elquesifilma]

I need to move 60,000 barrels of water from point A (elevation 0 ft) to point B(elevation 100 ft). The distance is 5000 ft. The line is fixed at 8.8 inch internal diameter.

A tank at point A feeds my pump which will pump the liquid to a tank at point B. Using Hazen-Williams equation i calculate that i need 123 psig at the discharge of the pump to get to point B with 15 psig (to get on top of the tank assumed 15 psig would be sufficient).

My questions are:
1) What safety margins do engineers use for pump discharge pressures? Do I quote a 123 psig discharge pump or a 150 psig?
2) Is there any concerns if I enter the tank at 30 or 40 psig if I oversize the pump? What pressures should I hope to see at the inlet of the tank?
3) my system will vary from 30-60 thousand Barrels of water per day, should I have 2 pumps each to handle 30 barrels, or just 1 to handle the 60, what happens to it when only 30 are being pumped through it?
4) Can someone give me advice in how to control my system? The water going into tank A comes from a pump upstream of it which can be shut down if tank level gets too high.

Well, hope people find this problem interesting; I appreciate any comments, THANKS in advance!!

 

[katmar]

1) Safety Margin It is hard to give a generic number for a safety margin. You need to link it to the accuracy of your input values. When I see round numbers like 100 ft elevation, 5,000 ft length and 60,000 barrel/day I suspect they are not very accurate. Remember that pipes can corrode and get rougher in the long term. Generally pressure drop calcs can be within 10%, but you need to know what eventualities you are catering for.

2) Pressure at destination This is one of the more common misconceptions in hydraulic calculations. You cannot have 30 or 40 psi "left over" when you reach the destination. The pressure at the inlet of the destination tank is fixed by the pressure in the vessel and the head of liquid above the inlet. If the pump head is more than your calculated head for the desired flow, all that will happen is that the flow will increase until the head delivered by the pump matches the head consumed by the system. Find out how to plot a system curve on a pump curve to determine the operating point.

3) Turndown This depends on your control and maintenance decisions. A common policy in this sort of situation would be to install 3 off 30 MBBL/d pumps so that one can be repaired while you have two operating.

4) Control This is too wide a subject to give a comprehensive answer to here. Find a tame control engineer and discuss this in depth. There are many alternatives. At the simple end you can use "bang-bang" or "on-off" control. For example, if you used 2 off 30 MBBL pumps you could switch off both when the level in tank A gets down to 10%, start one pump when the level gets up to 30% and the second when it gets up to 60%, when the level goes below 30% stop one pump.

At the sophisticated end of possibilities you could install a modulating valve with a full PID controller.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[Artisi]

is this a home work problem?

 

[BigInch]

The first configuration you study should be a pump-pipeline with a discharge pressure control valve at the end of the pipeline to tank.

15 psig is equivalent to a 35 ft height, presumedly to the HI-HI (pump shutdown) level of the tank.

Any excessive discharge pressure over 123 psig, will require a taller tank or a backpressure control valve to keep the discharge pressure at 123 or less.

Ask for the vendor to quote a pump with 285 ft head and a flowrate of 60,000 BPD @ BEP. When you decide on a basic pump model, impellers can be selected or adjusted to zero in on your target flowrates at best efficiency flow and head.

Note: that your velocity is about 10 fps, a bit high, so you need to check waterhammer surge pressures.

Note: I get less of a pressure drop using the Churchill equation, so if your equation is high, you'll need extra tank height or a discharge pressure control valve.

Of more concern is what happens at 30,000 BPD. Your flowing pressure loss is much less, so if you keep the 123 psig discharge pressure at the pump, you arrive at the tank with excessive pressure and will overflow the tank, if the pump is not shutdown. Alternatively, a discharge pressure control valve can adjust the discharge pressure accordingly to reach the top of the tank.

One or two pumps is sometimes an operating cost vs reliability decision. With 2 x the flow at max, it leads itself to a 2 pump configuration quite readily, but costs will generally be higher. Is there a situation where you should always be able to pump at least 30,000 BPD, when one pump is down? That's a good configuration decision point.

Another alternative may be one pump (or two) with either a control valve or a variable speed control, but that requires quite a bit more study, after you have reached a decision on the one or two pump issue, and if you can determine the real necessity to pump at different flowrates and, since it appears that you have a tank in the system, your tank will be able to fill and drain to accomodate changes in flowrates, so maybe you want to consider variable speed later.... or maybe not.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[chief]

Do not forget to put a vent/overflow on your tanks. To prevent excess pressure in the case of the high elevation tank and to prevent vacuum collapse in the low elevation tank.

Offshore Engineering&Design

 

[BigInch]

BTW at 10 fps, you're probably going to want a control valve somewhere in there to allow slow velocity changes when you turn the pump on or off.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[BigInch]

Getting back to the velocity "problem", if it is a problem. Since you say you have variable rates, you might try flowing more time at some intermediate flowrate and upping the minimum flow from say 30 to 40,000 BPD and reducing the max from 60 to 50 K.

Better yet, evaluate the reality of the 60K flowrate. If you can pump 24 hours at some constant flow like 45K, and let the tank handle overage and underages, easy operations and no waterhammer is worth a lot in itself.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[elquesifilma]

Thanks for the responses. This problem is not a homework problem, it is a project i should be implementing soon. I actually have been told we will be flowing about 40,000 bwpd thru the line, so my velocity will be about 6.15 fps and required pressure of 87.5 for the pumps to get from A to B. I decided on 3 pumps, 2-50% pumps and a spare. Some concerns that i have are:

1)Should i quote a discharge of 100 psig (15 % margin)? and have a discharge pressure control valve to adjust the discharge pressure to reach top of the tank.

2) if my flow drops down to 10,000 BWPD i might be at 1.54 fps, i will be concerned about depositing particles, is there anything else that should concern me from the low velocities? This might happen for a couple of days, and then goes back to 40,000BWPD

3) where does the discharge pressure control valve go, downstream of the pump close to the pump or tank?

thanks in advance!

 

[BigInch]

Usually the best place is between the end of the pipeline and the tank inlet, unless your pump discharge pressure is greater than the pipe allowable pressure, then you will need one there too.

Keep the velocity over 3 ft/sec to avoid deposition of sand, etc. 1.5 is too slow. The reason to operate so slow is not apparent. The system should be distributing from the tank during those times with the pump turned off. If for some reason there is NO way to avoid that, you will be running the pump at very poor efficiencies.

Two 50% pumps and a spare might be overkill unless you have an extremely critical need for supply, or maintenance is not available for a couple of months at a time. It also lends itself to either 30 or 60, and what you really need is perhaps something like 45 yo 50.

Is the tank constructed yet. If not, increase its size to cover demands when you would otherwise be at low flow conditions. No way do you want to run 10,000 BPD with that system you described. Its 30% of 1 pump flow. Too low.

You must coordinate your design with intended and efficient operations to a better level than you have now. Conceptual design isn't completed yet ... IMO, so don't order the pumps just yet.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[BigInch]

If you draw a straight line between pump and tank inlet, is there anywhere that the pipe will be higher than that straight line? If so, where and how much higher? Now, where is the lowest point on the pipe and how much lower than the straight line?

You don't know the discharge pressure or pipe pressure rating you need until you check surge. It could be 123, 175 or even 200 or more. Order the pump by flow and head, let them tell you what the maximum discharge pressure on their recommended unit can be.

The pump's shutoff head and pressure equivalent verses your pipe pressure rating should usually be the deciding factor in determining a need for a discharge pressure control valve.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[BigInch]

and...

With your static head, the pumps will operate down to 80% of BEP, if on variable speed control. Lower than that and there's not enough VSD head and system flow is at 1200 gpm (40,000 BPD). For 2, 30,000 BPD pumps on VSD Operation, that's a minimum of 24,000 BPD with one pump. If you need lower, VSD operation is out, or pumps must be smaller capacity. Going lower than that flow, sedimentation is possible.

Running constant speed, but with a lower minimum flow of 50% with 1 pump of 30,000 BOP capacity would be an absolute min flow of 15,000 BPD, but constant speed pumping cost/gallon could be quite high at that rate.

Kinda suggests that minimum flow should be 24-30,000, depending on if you think VSD is good for you or not.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[Artisi]

You have been given some good information which should assist in a better understanding of what you need to consider.
The only additional advise I can add is to look at the option for the varying flow rates, ie, 3 x 50% pumps with one on VFD, this might cover the varying H/Q conditions. However to make this a meaningful study with answers with which you can select equipment you need to undertake a full hydraulic study by plotting all the varying conditions across a selection of pump curves to decide on the best combination of pump/s.

 

[BigInch]

I think 3 is too many for this small system, but necessary if 10 KB is definately - absolutely - no way around it - required. A bigger tank would be better still.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[elquesifilma]

BigInch, Artisi and katmar, thanks for all your responses and recommendations. I will analyze and think about the system further, and get back to you guys if i have any other concerns or to let you know how it all works out

Thanks!