Best practices, refinery pipeline contamination

[rocketscientist]

Does anyone know of any best practices for lining up flow from a petroleum products tank to a pipeline for a pipeline that carries many products?

And, does anyone know of any criteria for ordering new pipeline valves to reduce the risk of cavity contamination from use with multiple products in pipeline service?

 

[BigInch]

Products should be shipped in a series that is designed to first, minimize interface generation volumes and second, to minimize the cost of lost product, (interface mix is blended into and downgraded to the lower value product) and 3.) to minimize the cost of reprocessing any interface mix that is not possible to process.

The volumes moving around in most refineries (in pipe where it is possible to mix products, ie it should not be possible to generate an explosion by mixing any two products that are possible to enter into the same pipe) would not justify worrying about such small volumes trapped in valves. If it is a problem, use double block and bleed valves and drain them between changes in product runs.

Independent events are seldomly independent.

 

[LittleInch]

I agree with all BI says, but not sure what you mean by "lining up flow". If you mean how to program which product follows which then this is called batching and is quite complex as you really need to try and maximise parcel volumes to reduce interface sizes as well as the issue of which follows which as BI says

If you mean physically then your main manifold needs to be as short as possible and lengthy of time two valves are open reduced to the minimum with fast actuators.

Normally the amount of product in a line valve is negligible in terms of contamination, but clearly you want to keep it to a minimum. I don't know of any criteria, but you could ask vendors and then compare different valves.

A few more details and you'll get a lot more info.

LI

My motto: Learn something new every day

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

 

[rocketscientist]

Actually, I was looking for industrial standards for contamination elimination.

What I did find was:

API-570, 5.3.2 Dead Legs.
"The corrosion rate in deadlegs can vary significantly from adjacent active pipe." "Consideration should be given to removing deadlegs...."

API-2610, 10.3.5 Dead Legs.
" .... the construction of dead legs shall be avoided and reduction or elimination of idle pipe sections should be considered during design and construction."

API-6D (new version)
{Double-Isolation Block (DIB) preferred over Double Block and Bleed} A DIB is a "single valve with two seating surfaces each of which in the closed position provides a seal against presure from a single source with a means of venting/bleeding the cavity between the seating surfaces." Looked up an article called, "The true meaning of double block and bleed."

I did some searches in the area of jet fuel since their standards are strict. I found nothing on pipe layout. I am currently talking to several valve and pipeline people regarding standards for layout, lining up flow from a product tank to a pipeline and valve selection.

 

[LittleInch]

There is very little in "standards" other than company guidelines and good practice. That's kind of what I meant by saying your initial manifold should be a short as possible and clearly any dead legs on the system from that point onwards should be eliminated or reduced to a minimum. There was always an issue about what you did with the pig traps if you had a pig trap full of one product, but didn't want to drain it away.

Basically, keep it as a single pipe from the tank without tees, branches etc unti you get ot the manifold where you want everything (valves, tees etc) as close together as you can to reduce dead legs. Aviation should be at the back of the manifold so that it sweeps all other liquids before it, thus reducing the potential for contamination. Aviation needs separate inlet and outlet piping and you can't flow in while you're flowing out - you can for diesel(gas-oil) and petrol(gasoline).

I have some fuel piping guidelines squirreled away somewhere (on paper), but the above is the essence of it.

What you need to do is look closely at the actual piping layounts and designs when they are produced, because lots of piping engineers will not see the little note on the P & IDs which says "dead legs to be avoided / kept to a minimum distance"... The company I used to work for who ship lots of aviation and multi product, always used to use Orbit Truseal valves - see

http://www.c-a-m.com/forms/Product.aspx?prodID=41758ecc-7e41-46ab-9adb-44dd00000bbe

or search for them. Great isolation valves which could also auto bleed as part of the closure mechanism and alarm if product kept flowing.

My motto: Learn something new every day

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

 

[rocketscientist]

Funny, I was just looking at that brochure at the beginning of the week. The flaw I see in that type of valve is that the vents should be blow out every time a product is changed to avoid contamination. Even then there will be problems. When I worked in Pharma and foods, butterfly valves were required because they could be easily cleaned. Of course, they make for lousy shutoff valves. Again, we had no standards for pipe design we merely had a set of principles and built from them.

 

[BigInch]

So "lining up flow" means the piping and valve configuration???
That usually refers to the sequence of batches being pumped into a multiple product pipeline.. where I'm from.

Independent events are seldomly independent.

 

[LittleInch]

That's what I thought it was about, but could have been either. Rocket scientist, I think you're creating problems with the valves where none exist. The isolation valves quoted are normally to isolate one product from the multi product line or a meter run and the valve drain is only to demonstrate sealing, i.e. it doesn't drain the whole valve, but if there is any leakage from either seat, allows a path out rather than through the valve.

One way of checking a piping run is to do a markup with each of the different products on a separate page from tank to delivery point / pipeline and see where the dead legs lie. Look at things like pumps and meter runs which are not always in use to see where potential contamination volumes exist. I really wouldn't bother about valves though.

My motto: Learn something new every day

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

 

[BigInch]

Well I've also heard it used as setting the positions of various valves in a large system so that the correct product flows into it from the correct tank, but I've never heard it being used in the context of designing a piping configuration.

Independent events are seldomly independent.