block valves at river crossings

[pipeliner3000]

ASME B31.4 requires isolation valves to be installed at major river crossings. ASME B31.8 does not. We have a multi phase trunkline collecting well fluids that are mostly liquid which crosses a river (small channel 5m but large flood plain upto 500m flooding in wet season). As it's multi phase the code selection is gas ASME B31.8.

- by the code we don't need to install isolation valves at the river. However, as its mostly liquid and the flood plain is quite sensitive should we be considering valves? how do you determine a 'major' river. Do these need to be automated/ with leak detection to make them have any impact? Don't see manual valves actually reducing environmental impact that much as the horse would have bolted before anyone got there in time to close them. Also don't want a complex valve control / leak detection system.

 

[zdas04]

First off remember that the code is the minimum, not the maximum. The code (31.8) does not require PSV for thermal relief on a pig launcher/receiver, but it is prudent to add one. Same way with the river/railroad/interstate crossings.

Think about the reasons for the block valves. If a line is single phase gas and an anchor drag splits the line in two, what happens? The line blows gas into the water and the chance of fire/explosion/ground contamination is nearly zero (fire is a tough concept since there is no way for it to get back to the source).

On the other hand, if the line has hydrocarbon liquids in it then that same anchor drag will drain the line into the river and could easily catch fire, contaminate the river banks, and generally be a bad thing. If your "gas line" has more than dimenimus liquids then I'd put the block valves.

As to size of the river, the EPA is now claiming that ephemeral washes (that may only run in a 100 year flood) are Waters of the US. Historically, we drew the minor/major line at the Waters of the US designation. That will soon no longer be possible. I would define major as "expected to flow water during an average dry season".

How do you automate the line? Well, once you get a block valve installed you have a lot of options. Which one you pick really depends on your company's risk tolerance. You'll likely be doing a HazOp on the line and the level of instrumentation/control should come out of that.

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

 

[bimr]

ASME B31.4 434.15.2 (a) requires mainline block valves on the upstream side of major river crossings and public water supply reservoirs, and either a block valve or a check valve on the downstream side. 49 CFR Part 195, ―Transportation of Hazardous Liquids by Pipelines,‖ has incorporated ASME/ANSI B31.4 code by reference.

On critical water crossings, some pipeline operators are installing a remotely operated isolation valve on the upstream side of the river crossing and a check valve on the downstream side of the river crossing.

 

[zdas04]

The problem I've always had with the check valve downstream was the marketing kept finding new markets that required trunk lines to reverse flow. You really don't save enough money to be worth blowing down the whole line to replace the check with a block valve if that happens.

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]

I'm intrigued by the phrase "As it's multi phase the code selection is gas ASME B31.8."

If you look in the code at the start, 802.1 b) "This code does not apply to:... 6) ...flowlines between wellhead and trap or separator.."

Also 802.2.1"The requirements of this Code are adequate for safety under conditions usually encountered in the gas industry. "

Note "gas industry", not upstream flowlines.

Therefore if it all goes horribly wrong and your design of pipeline based purely on this code is questioned, how will you answer this question (not in the scope) in court?? There's not much wriggle room there or so some lawyer who only works in black and white will tell you. I'm well aware that in practice many flowlines are designed to either 31.8 or 31.4, but that won't wash if it goes wrong.

Therefore given the deficiencies which occur when using a design code intended for one thing (dry, liquid free natural gas) for another (multiphase mixture of hydrocarbons and other gases) you need to think outside the "code requirements" to demonstrate that you have considered all the potentials issues and addressed them properly. If a similar code (e.g. B 31.4) has more onerous requirements, then it would seem a good idea to follow them, noting also that B 31.4 doesn't cover flowlines either, but at least you could claim to have taken the more onerous conditions from either code or another code which does cover flowlines.

Whether you need valves is a risk issue with environmental risk being a key part. How big / long your line is, what the profile is, what your systems are for discovering leaks (multi phase is notoriously difficult to do with instruments alone) and how sensitive this river / waterway is your decision and in your head, but remember that any leak would continue to leak until the gas pressure dissipates and a valve, even a manual one, can reduce the timescale and volume of a leak quite considerably.

Realistically a 5m wide stream will not have vessel traffic and a corrosion leak (film of oil etc) is your most likely event, where manual valves would be adequate to stop the flow whilst you repair or replace this section. Auto valves would seem a bit excessive unless there is some environmental sensitivity.

With your wide flood plain though it may be you need to put the valves at the edge of this so you can get to them if you need to and also protect this area during the flooding season.

Given this is a flowline it is unlikely to be needing to be reverse flowed, but remember many pigs don't really like check valves and an isolation valve is usually better.


My motto: Learn something new every day

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

 

[BigInch]

B31.8 is often selected (rightly or wrongly and despite the stated scopes of codes) when using Shell PDs (require B31.8 when Vapor Press at operating temperature is > atmospheric) and some other companies philosophies, probably because of .8's use of area class design factors, which although are less conservative when out on the pipeline, actually require a thicker walled pipe inside stations than will B31.3.

you must get smarter than the software you're using.