Pressure testing pipelines with ULG or Diesel 1

[321sol]

Hello,
I have to pressure test some underground steel pipelines using the fluid they carry, i.e. diesel or gasoline. A previous test showed that the pressure dropped about 1.5 bars from 12 midday to 9AM the next morning. I need to determine whether that pressure drop is due to the thermal of the fuel and any air pockets or due to a hole.

Tests prior to the one mentioned above recorded high pressure drops in shorter periods of time. There is a possibility air was present in the line during those tests. No measurement of temperature was done but variations can be to as much as 5 deg. Celsius.

I need a standard test procedure to redo this pressure test. Is there any blank flange commercially available with possibly a thermocouple or an RTD to measure the temperature of the fuel inside the line?

 

[zdas04]

You can buy a threaded blind or have a welder cut a hole and weld in a half coupling (this is my preference since I've had problems with NPT threads in a blind from the factory being too loose for a test). I've probably had 500 of these "threaded blinds" made over the years and they don't cost much while being very effective.

With a water test my procedure call for:
[ol 1]
[li]Fill the system to the vents, let it rest for 2-3 hours[/li]
[li]Refill it to the vents (and if it takes more than about 5% of the pipe volume I give it another rest period)[/li]
[li]Shut the vents and raise the pressure to somewhere around 30 psig[/li]
[li]Let it sit over night with pressure on it (final pressure must be greater than 0 psig, I generally have someone check on it every 4-6 hours and add water to get it back to 30 psig) then continue otherwise start looking for a leak[/li]
[li]Open the vents and fill the line again (if it takes more than about 2% of the volume then repeat steps 1-4 [/li]
[li]When the system passes 1-5 then start the test[/li]
[/ol]
I know this sounds like a royal pain, but with lines that have a volume that is more than 40 bbl (say 1,000 ft of 6 inch), not doing this is a bigger pain.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[LittleInch]

One of the issues with using hydrocarbons is they expand and contract quite a lot.

This FAQ can be used in reverse to see how much preassure you can lose with only a few degrees loss.

http://www.eng-tips.com/faqs.cfm?fid=1339

The temperature you really want/need to know is the 95% or mitre which is underground. To get that you need to digg a couple of holes, attach surface temp probes and then backfill

Dave has some very useful guidance.

Get the temperature steady or do it for at least 24 hours to get a compete day cycle

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

 

[321sol]

Thanks LittleInch and zDas04.

zDas04, I would not be using water in my test but I can use your method of letting the fuel sit for a period before beginning the test. The problem I foresee is that a small amount of the fuel may vapourise in the underground section and that section is fairly horizontal which means that there will be cavities of trapped vapour. Also, it is possible that there may be cavities of trapped air after the lines are filled with fuel.

LittleInch, I am going to review that link and get back to you. Finding the isothermal compressibilities for petroleum has been difficult. I found some tables at

http://nvlpubs.nist.gov/nistpubs/jres/5/jresv5n5p985_A2b.pdf

but nothing anywhere else to verify the values.

 

[321sol]

LittleInch, your link seems to be exactly what I need to do the calculations but I need the compressibility factors for petroleum fuels i.e. diesel and gasoline. Where can I find those?

 

[zdas04]

321sol,
We do a lot of tests with water. There are parts of the learnings from those tests that apply to other fluids. I was just trying to get you to think about the parts we've already learned so you can capitalize on it. The soak at low pressure helps with the gases evolving out out the liquid.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[321sol]

David, just confirm for me that the lines are gas freed before you pressure test with water. In my case, where the lines are shaped like an widened 'U', and the fact that they would still be primed when disconnected, should I go through the trouble of emptying them and refilling with water? Wouldn't it be easier to just use the same fuel to apply the pressure? Wouldn't doing this also reduce the possibility of air in the lines? My only trouble with this is finding the compressibility factor of the fuel.

 

[BigInch]

don't look for compressibility of diesel, google "bulk modulus" instead.

 

[zdas04]

321Sol,
NO. I am giving you learnings from hydrotests with water. Many of those learnings translate exactly to a static test with other liquids. If I'm testing with diesel (and I have), I fill it to the vents. Let it sit for a while at atmospheric pressure to let any vapors that are going to evolve out to do so. Bring it to a low pressure and let it sit to let any vapors willing to condense to do so. Stabilizing the fluid for an equilibrated ground (or ambient) temperature is a good idea with any fluid.

As BigInch implies, Bulk Modulus is the reciprocal of compressibility.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[321sol]

Thanks everyone, you have all been a tremendous help. I'll let you know how it goes.

 

[LittleInch]

Just be sure that you have a high confidence in the ability of the pipelines to withstand the hydrotest pressure without failure.

Testing in hydrocarbons always carries a certain risk of failure.

In most cases that failure level is acceptable, but nowadays you will need to document why you believe it is acceptably low.

I fully understand that it is seen as a good idea to proof test pipelines every now and then ( I think the operating company I used to work for did it every year), but in the main this type of test has been overtaken by risk based criteria and use of internal inspection tools and external coating inspection techniques.

Make sure you have a plan in place and primed in case any of the lines go POP....

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

 

[321sol]

David, have you ever used deaerated water or fuels? I just found an interesting graph that shows the effect entrained air has on the pressure loss.

Littleinch, thanks for the heads-up. I am slowly gravitating towards using water as the fluid to apply pressure.

 

[zdas04]

I never have. I looked into doing that for a project once, but the cost wasn't worth the benefit on that project. Never looked at it again.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[321sol]

Check

http://theleeco.com/engineering/liquids/bulk-modulus.cfm

3.5% entrained air for a fluid having a working pressure of 100psig can have its bulk modulus reduced by 90%.

 

[BigInch]

Which is exactly why it is very, very imporatant to allow enough settling time to let that entrained air dis-entrain itself and be flushed out a vent before beginning the test.

 

[321sol]

A vacuum pump hooked up to the line can get rid of a lot of that entrained air in no time. After which the vacuum pump can be disconnected and the pressure pump connected.

 

[zdas04]

Good luck with that. It can work, but people generally turn the pump off too soon to finish the job. Then bad things happen. You never said how long your line is, but if it is long enough to matter then you'll be way better off waiting for the temperature to equilibrate and the gases to shift around on their own than running a vacuum pump until you see liquid in the suction and calling it good for your test start. You sound like someone who will have to learn that lesson for yourself, so try it. Let us know that we're all wrong and your idea was the right way to go all along.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[LittleInch]

But where's this air coming from? 3.5% sounds a kit to me for a system which starts from a tank and doesn't involve sprays, free flow or low pressure locations.

A vacuum pump will only work in something like a dearator tower. Not normally required.

Let us know how it goes.

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

 

[321sol]

No need for the sarcasm, David. The line is about 30ft. long. By waiting, what percentage of entrained air would leave the fluid? And under what mechanism would it do so?
In my estimation, lowering the pressure above the surface of the liquid (not less than the liquid's vapour pressure of course) will remove a lot of air. One would not see the liquid in the suction line because it is imperative that all connections are sealed and the piping to be rigid. After which, the pipeline will be pressureised and left to allow the temperature to equilibrate. The test will begin about an hour or two thereafter. All the vacuum pump is used for is to increase the rate at which the gases escape by lowering the pressure above the liquid surface.

Littleinch I would most likely be using water as the working fluid and so it is possible that there would be some amount of entrained air present.

 

[LittleInch]

OK, then don't entrain air is the first point.

The second you need to think about is the surface area issue. The surface area available to you at any one point in a piping system is minute compared to the volume. Thus a small negative pressure applied at one of a piping system - it's stretching things to call 10m of pipe a "pipeline" BTW - the amount of any entrained air coming out which wouldn't come out anyway is negligible.

There's a good reason why vacuum de-aerators work on a column principle where streams of water fall vastly increasing the surface area.

I think what Dave is saying is that this hasn't been a notable problem before so I think you're in danger of overthinking this, especially for such a short bit of pipe.

Filling it in such as way as to drive the air out is usually the issue - using a pig for 10m of pipe is overkill, but not sure whether your line is flat, sloping, buried or indeed anything about it other than what you've told us.

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