Riders on the Storm: Stopping and Restarting Offshore Oil and Gas

It has been five days since Hurricane Gustav blew through town, and industry is still working to restore the flow of oil and gas from offshore production platforms in the Gulf of Mexico:

Meanwhile, about 47 percent of more than 700 stationary offshore platforms in the Gulf of Mexico remained evacuated Friday, according to the Minerals Management Service in New Orleans. The agency also said that 34 of 121 oil rigs remained unmanned.

Under the fold the latest from the industry on the LOOP, Port Fourchon, and other infrastructural concerns.

Likewise, the LOOP (Louisiana Offshore Oil Port) has reopened. However, it is running on backup power such that only one tanker at a time can be drained of oil. Commercial power might not be restored to the offshore terminal for another week or two.

With Hurricane Ike possibly headed for the same area in about five days, one might consider whether it is worth the effort to go back out there and restart things until the latest storm passes. What is the process for shutting down production and evacuating prior to the storm? What (and how long) does it take to do the reverse?

This document (1 MB pdf) from the National Ocean Industries Association provides some answers to these questions and well as information on regulations for platform design and operations with hurricanes in mind. For example, here is the shut-down procedure in brief:

  • Evacuation Phase 1 - Receive Storm Notification
    • Review operations forecast.
    • Communicate with air and marine transportation
    • Perform safety system checks.
  • Evacuation Phase 2 - Complete Preparations
    • Secure all equipment.
    • Test communications systems that enable monitoring
      from shore.
    • Evacuate non-essential personnel.
  • Evacuation Phase 3 - Shut Down and Evacuation
    • Shut-in wells and subsurface safety valves
    • Close incoming and exit pipelines
    • Shut down operating systems
    • Transport remaining personnel to shore

And here is the start-up:

  • Visual Inspection of Affected Area: Operators perform fly-by
    inspections of facilities for preliminary assessment of damage.
  • Pre-Boarding Safety Meeting: Facility personnel and managers
    meet to review post-storm inspection information.
  • Damage Assessment: Core personnel return to the facility to
    conduct a more thorough review, which may be complemented by
    data underwater ROVs (Remotely-Operated Vehicles).
  • Equipment Integrity Verification: Review of the physical topside
    structure and the pipelines is completed.
  • Safety System Function Test: Individual systems are brought
    back on-line to be tested one at a time.
  • Facility Startup.

Once shut down, the biggest problem is with damaged pipelines on the sea floor. These can often be repaired or bypassed without too much delay, but damage to the platform itself is a much bigger headache. The poster child for a platform suffering from a bad air day is the Thunder Horse platform, which was left atilt by Hurricane Dennis in 2005.

Shell's Mars Platform stayed upright during Hurricane Katrina, but it took nine months to restore the oil flow -- no small feat given the photo below:

At the time that the above-linked NOIA pdf was written, 343,000 BPD of production was still shut in from platforms damaged by hurricanes the previous summer (2005).

A more politically-sensitive issue is environmental risks from hurricane damage, and the NOIA document highlights the various technologies -- mandated by regulations -- which prevent the spillage of oil from ruptured pipelines.

All offshore platforms are equipped with safety valves that shut-in
oil and natural gas in the event of storm damage. These valves lock
closed at regular intervals so that oil or gas cannot flow if equipment
is broken or separated. Every single safety valve held during the
2005 hurricane season.

Most of the damage to platforms has occurred on structures built prior to the 1988 regulations. I don't know, however, to what extent that older infrastructure has been retrofitted with equipment designed to prevent spills. Given the current volume level of the offshore drilling issue during the 2008 US Presidential election, an incident of any consequence due to Ike (or perhaps Kyle) will easily ratchet the noise up to a level of eleven.

I think it is great that the oil companies do shut these things up tight for storms. My guess would have been that they run full tilt and sacrafice the environment and their own worker to get one more barrel out.

Maybe we can all agree that safety and the environment are more important that oil.

Does the reservoir pressure build up again while the plugs are installed? Are flow rates greater for a while afterwards?

I can't resist:

Of course the reservoir pressures and flow rates are greater after a shut in.

They go up to eleven.

(See link at end of key post.)

It is actually why you will see credit given on occasion to techniques that don't realistically work. Since almost every time a well is shut in the production goes up initially when the well is re-opened, it gives possible credence to whatever "black box" treatment has been used to improve production during the shut-down. It can take a while to show that the "magic bullet" didn't actually do anything.

The most common measure of effectiveness for well interventions is 'skin' (SLB Oilfield Glossary defines it as the zone of reduced or enhanced permeability around a wellbore, often explained by formation damage and mud-filtrate invasion during drilling or perforating, or by well stimulation). Simply shutting in the well for some time and then re-opening will temporarily increase pressure and flow rates, but it will not affect skin. With a successful well intervention, the calculated skin will go down and the well will be more productive.

I gotta watch that movie again.


It's a fine line between clever and stupid.

Maybe we can all agree that safety and the environment are more important that oil.

Ha! If we all agreed on that, the world would be quite a different place, wouldn't it?

I like Ike... But he hasn't been running for office and this Ike seems to be crawling around.

I am very worried about the slow track over Cuba.

It seem to indicate an extreme variability in the storm track.

It looks like this storm is just going to dance around the Gulf picking up strength, rotation, humidity, and might strike land, eventually, as an absolute monster.

If it heads north to "N' orleenz" its going to be really really BAD.

It doesn't have to hit the city. Just take out a part of the platforms and LOOP (even Texas, south of Houston, would be very damaged by a cat 5.)

New Orleans is barely recovering from Gustave. gasoline must be very precious right now over there. As someone mentioned before, they might not have enough gasoline for the people to evacuate again. let alone their personal bank accounts to finance another hurricane vacation getaway.
on a seperate note, models are suggesting Houston, TX. (generally speaking, but still a big possibility) If Houston get a direct hit,
it will be very ugly, worse than Katrina was for New Orleans.

Remember the mass exodus from Houston when Rita was threatening? bumper to bumper traffic all the way from Houston to Dallas. and thats a 4 hour drive at posted speed limits. it was 24 hour drive minimum at least for everyone for a while. only to find there are no motels.

but many folks i have spoken to, do not plan to leave again for that last memorable experience. if they really do stay, and Houston gets slammed with a cat 4 min, it will be ugly. models are showing no higher than cat 2, maybe on the high side of cat 2 and close to cat 3, but we all know things can change.

Then there is the damage asessments are for the oil producing infrastructures.

only time will tell!

It seems after a brief spike, gas prices in New Orleans only went up to national average (so far)?!


(add the national average to the chart manually - it seems I can't link to the chart with the average included).

No shortages and prices appear to be dropping in New Orleans proper.

Elsewhere, slowly getting caught up in blackout areas as power comes back on generator use declines.

Lake Charles evacuated for Gustav and seems likely to need to evac again.


With each model track update:


the models are pulling into Louisiana.

Frederic, 1979 into Mobile and Elena, 1985 meandering are the most similar
to Ike, and neither of these cut thru to the Western shore of Cuba as fast as
Ike is.



HWRF and GFDL models, the LBAR and NOGAPS are now turning into LA.

I'm convinced the models do not take into account previous storms,
such as Edouard, Gustav, and Fay.

Those "ghost tracks" will move Ike toward Mississippi/Mobile.

Either way, a Cat 3 will be 250 miles S of NOLA and crude is at $108.

When there is zero predictive value in a model, what good is it?

Off topic JB but thanks for your response to my questions the other day in the DB.