Is a brownfield hydrotest safe on a late-life asset? 2

[dean427]

Hi all.

I have run a quick search on the forum be couldn't find anything that covered this topic so hence the question.

Caveat...I'm a process engineer rather than mechanical so I apologize in advance for any ignorance below.

The question is...Is it safe to complete a hydrotest on an aging/"late-life" operating facility? Facility is an onshore hydrocarbon gas conditioning and processing terminal.

Background:

We have identified via a HAZOP review that a gas compressor's suction pipework may have its design pressure exceeded during a settle-out case when the compressor trips.

My Mechanical Engineering Technical Authority wishes to hydrotest this section of pipework in order to re-rate it.

I trust his experience (much more years under his belt than me) however I feel that a hydrotest being completed on a circa 40 year old gas plant is very risky - i.e. how can we assure ourselves that every mm2 of pipework has been inspected to ensure no defect which could cause the pipework to fail during the test at extremely high pressure? Not to mention trying to dry the system after the test should it be successful will be very difficult!

For info...gas plant has x2 processing modules. The intent is to complete the hydrotest during a turnaround/shutdown on one module with the second module still running (other workscopes will also be ongoing on the shutdown module) & site will be fully manned.

Is this something that is fairly routine in wider industry? My technical authority seems to be driven to do this to ensure code compliance but I feel that purposefully exposing a system to the hazard you are trying to mitigate against is not logical or sensible.

Grateful in advance for any advice or steer regarding this issue.

Thanks

Dean

 

[georgeverghese]

Settleout pressure should be accomodated within the specified mechanical-process design pressure limits of the affected piping and vessels. I've never come across a case where compressor settleout condition is accomodated within the hydrotest pressure limits using the "loophole" in B31.3 which allows for rare cases where the upset case pressure falls within the hydrotest pressure limits.

You say the settleout pressure MAY exceed the design pressure limits - can you explain ?

 

[mk3223]

IMO, it's required to have a hydro test for the system, either to be re-rated to higher or lower pressure ratings, to ensure the system can be operated at a pressure which is intended to be, in spite of the age of the equipment. So, you may identify some items to be replaced or upgraded as necessary for the continuous safe operation.

 

[EdStainless]

Is any other NDT planned? ID inspection? PT of welds at joints? Usually based on maintenance history.
Normally some critical points would be checked and then the hydro.
You can't run the plant if you can't hydro it
If you are worried about it passing the test then it should be scrap.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[dean427]

@georgeverghese - yes suction pipework design pressure can be exceeded in certain cases as identified by the HAZOP i.e. max mol wt case with MOP at the suction side and max feasible head across the machine. This has not happened in the past but it is feasible and credible.

@EdStainless - we have our current inspection records to go on here. It is certainly feasible to request a more (and possibly more detailed) inspections which I think is prudent. You say you "can't run the plant if you can't hydro it" just to be clear - the pipework has been hydrotested originally when the facility was constructed/commissioned - the new hydrotest would be to re-rate to a higher design pressure now - this will then be considered an engineered solution to the overpressure case as described above. Just because we have identified a hazard in a retrospective HAZOP does not mean the plant is unsafe to run currently.

I would like to know at least...is it a code requirement to hydrotest for a re-rate or could this be done by analysis?

 

[LittleInch]

dean427,

Appreciate your issue and to a certain extent you're in no mans land here - anything you do is likely to have consequences.

To me I find it difficult to understand why anyone feels the need to re-rate the design pressure of a system that has clearly stood the test of time at 40 yrs old unless you're changing something. There are many other ways I feel you need to look at first to either prevent this happening or mitigating it ( relief valves etc) or a George says using up the generous over pressure allowance in B 31.3, assuming that's your design code?

But your main question "Is it safe" - we can't judge. There are some 40year old plants that in terms of pipe integrity are as good now as they were 40 years ago, bar a bit of fatigue use. There are 10 year old plants I wouldn't work in because they are so decrepit. Which is yours?

Testing anything comes with the risk of failure. Adding water to a system which normally has gas in adds a LOT of weight and will be a bitch to de-water and dry. Maybe the pipework is OK, but the structure collapses.

Hydrotest implies water. Generally this is a lot safer than gas pressure so it's as safe as you can make it. Nothing is risk free.

It's up to you and your team to assess the risks from doing it or not doing it and see which one wins. Uprating by design and inspection can be done, but a hydrotest is relatively simple in theory and proves the point to everyone. It is though a pretty blunt tool.

Is it routine? I would say no because you often can't uprate something without replacing it or large parts of it and then that's a different testing story. I've not come across many instances where by undertaking a higher hydrotest pressure on something this old, you can successfully uprate the design / MAOP.

But the devil is in the many details only you have (e.g. what sort of pressure class are we talking about / change in pressure?)

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

 

[GD2]

dean427,
I will talk about what the code says assuming the facility construction code is B31.3.
From Code perspective, code allows over pressure as long as certain conditions are met per para 302.2.4. I quote below few of it:
(d) The total number of pressure-temperature variations above the design conditions shall not exceed 1000 during the life of the piping system.
(e) In no case, shall the increased pressure exceed the test pressure used in the original design per para 345.
(1) Subject to Owner's approval, it is permissible to exceed the pressure rating or the allowable stress for pressure design at the temperature of the increased condition by not more than
(a) 33% for no more than 10h at any one time and no more than 100h/y, or
(b)20% for no more than 50h at any one time and no more than 500h/y.

The code uses the use-fraction sum rule to give these time-dependent limits. I am not sure about your competency but these can be easily calculated if a good log is maintained. Be aware that the allowances are on the basic allowable stress used from Table A-1 of the code and NOT the design pressure. For flanges, the allowance will be on the B16.5 MAWP rating table. Again be aware that the allowances are not applicable to all type of material. Carbon steels are acceptable. Your ultimate finding should be that the increase in the pressure DO NOT cross the Yield Strength at temperature.

Regardless of the above situation, if your plant is 40 yrs old, it stand at risk if it was built to B31.3. Plant designed to B31.3 usually have a life of 20 -30 yrs. Plant designed to B31.1 are usually designed to 40 yrs. life.

Your plant must have a good integrity program, to make sure that the piping wall thickness do not fall below the design thickness at corroded condition. If it has, the plant will always pose a risk to fail even at lower pressure than design pressure.

I am not sure, why you will think doing a hydrotest is unsafe. As LI has pointed out, Hydrotest is usual in a shutdown/turnaround and poses lesser risk than pneumatic . But before you do, make sure that piping supports were designed to take the water load, otherwise you might have to provide extra temporary supports.



Ganga D. Deka, P. Eng
Canada

 

[dean427]

@LittleInch - Thanks for the detailed response. I feel that we are on the same page & agree that there are other options worth looking at. Hydrotesting this section of plant is the last thing I want to do. Thanks for the prompt regarding the weight - will have our Civil/Sturctural Engineer review the potential plans.

@GD2 - Thanks for the response detailing what the code states - this is what I was looking for. I fully appreciate your comment about hydrotest not being as risky as pneumatic which would carry the risk of rapidly expanding nitrogen gas cloud on release but for me it is the scale of the pressure rise required by hydrotest. Our company carries out pneumatic pressure reinstatement testing at 90% of RV set point - so a hydrotest at 1.5xdesign pressure is an order of magnitude above this and to me carries a greater risk of pipework failure. In addition, introducing water to a gas process where it will be subsequently difficult to drain and dry also carries risks - this could cause hydrate formation on restart leading to blocked outlet overpressure scenarios, collapsed or damaged strainers, compressor blade failure, plugged RV inlet pipes etc etc. All of this I would like to avoid if permissible.


Thank you all for the responses, greatly appreciated!

 

[georgeverghese]

Can we see a PID of this compressor unit which shows the suction piping / recycle valve tie in , shutdown and check valves - with piping classes shown on line numbers ?

In most plants, the settleout pressure is prevented from surging further upstream of the compression unit by a check valve placed at the suction side ESDV, and this check valve should be located upstream of all compression unit recycle line tie ins.

 

[dean427]

Sorry George but I'm not permitted to share the P&ID due to company information disclosure guidelines.

Machine is a 2 stage centrifugal compressor with a single "long recycle" line (i.e. 1 recycle valve for both compressors via the coolers) and a "hot gas bypass"/quick recycle.

The suction pipework in question is from the NRV right up to the compressor casing.

 

[georgeverghese]

Okay, so this settle out overpressure is within the compressor unit. How then are you dealing with this overpressure for the suction scrubber - ASME BPVC doesnt allow pressure exceedence beyond 10% of mechanical design ?

 

[Gator]

OT: LittleInch, do you know what happened to BigInch? He hasn't been around for about a year, apparently.

I used to have his email address but lost it somewhere along the line.

 

[LittleInch]

Gator:

BI is alive and well but has bowed out of ET for the moment after several years. He's currently doing up a house on a Spanish Island with his wife although I need to contact him again to see how it's going.

I'm not going to release contact details in a public forum.

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

 

[Gator]

No problem, good to know, was just curious. I looked around to find an E-T "Contact Member" feature for about 14 seconds and then gave up.

 

[dean427]

Hi George - there is no suction scrubber per-se on this system...the upstream "scrubber" is a demethaniser column. Dry sales spec gas from the top of the demethaniser is boosted via recompressors driven by turbo-expanders and warmed through the gas-gas exchangers. This upstream system is protected by multiple ESD valves as well as NRVs from the sales-gas compressors so is not at threat from the high settle out pressure. The column etc is also protected by adequately sized RVs.

 

[georgeverghese]

Okay, am familiar with the process systems arrangement on demethaniser overheads - so this settleout pressure concern is within the expander recompressor circuit only, and this settlout wave doesnt travel back through to the gas - gas exchangers and DeC1, since as you say, there is a check valve and SDV just upstream of the capacity recycle and hot gas bypass return line tieins at the main suction line.

It is not clear at the moment which settleout case exceeds the line design pressure - usually the worst case settleout would be based on max normal suction pressure from the DeC1, and max normal discharge pressure, which could be as high as this recompressor discharge PAHH. Another less severe, non controlling case would of course be normal suction pressure and normal discharge pressure. Which case is breaching your current piping mechanical design limit ?

I know of some existing compression systems which have similar problems, and they have made some awkward arrangments to decrease the settleout pressure, such as (a) to blowdown the compressor every time it trips, with the existing BDV on compressor discharge (b) adding a blow off PIC/ PCV to flare on the compressor suction line.