BP / Gulf Oil Spill – How Big? Just Got Bigger. Again.

Well, we warned ya. That estimate didn’t last very long.

Yesterday the government released new estimates of how much oil was spewing on a daily basis from BP’s Macondo well into the Gulf of Mexico, based on much better data and hi-def video. It’s worse than we thought: initially the well was gushing oil at a rate of 2.6 million gallons (62,000 barrels) per day. At the time, BP and the Coast Guard claimed the spill rate was 62 times smaller than that — 42,000 gallons (1,000 barrels) per day.

Eight days into the spill, NOAA got involved and raised the estimate to 210,000 gallons (5,000 barrels) per day. BP grudgingly accepted this, and it remained the official estimate of the spill rate – more than 10 times too small, we now know – until May 27, more than a month after the disaster began.

Working with BP as a consultant back in the 1990s, I interacted with some of the smartest geologists and reservoir engineers I’ve known. How could they get this so very wrong, knowing the bottomhole pressure of the well, the size of the reservoir, and the relative percentage of oil to natural gas? And how could the US Coast Guard – the agency that responds to oil and hazardous materials spills of all kinds and sizes – not know that they were dealing with a spill 10 times larger than the official estimate?

As the well leaked oil and natural gas, pressure in the reservoir below gradually lessened, so the flow rate of oil declined to 2.2 million gallons (53,000 barrels) per day. The well was capped on July 15, totally shutting off the flow of oil and gas.

Bottom line: using these new government numbers, and the estimate in a previous Washington Post article that about 33.6 million gallons were diverted from the leaking well and never entered the Gulf, we come up with a total spill of 172.2 million gallons (4.1 million barrels) – more than 15 times the official size of the Exxon Valdez oil spill in 1989.

Michigan Pipeline Spill – A Warning Shot

Turns out the pipeline that failed in Michigan last week, spilling a million gallons of oil into the Kalamazoo River that’s still threatening to reach Lake Michigan, was installed in 1969. Still no word on why it failed but corrosion is a constant battle for pipeline operators like Enbridge.

Like who?

Most oil, gas, and refined-products pipeline, onshore and offshore, is owned and operated by companies you’ve probably never heard of (without the brand name or deep pockets of a company like BP or Exxon). Enbridge is a Canadian company that claims to operate the world’s largest pipeline network – 15,000 miles of pipe in the US and Canada. They also have a rap sheet of recent, major spills and fatal incidents (although we don’t know if their record is any worse than most other pipeline operators).

Active oil and gas pipelines in the US Gulf of Mexico. Data from US Minerals Management Service (downloaded March 25, 2009).

There are about 25,000 miles of active oil and gas pipeline on the seafloor in the Gulf of Mexico, connecting 3600 platforms and tens of thousands of wells to coastal storage, processing and distribution facilities. Much of this infrastructure is getting old – drilling began offshore in the Gulf in the 1940s. Are pipeline operators doing a better job inspecting, maintaining and replacing the pipes offshore than they are onshore?

Keeping an independent eye on this vast, aging infrastructure is yet another reason we think Gulf-wide satellite monitoring should be a routine activity, not a service limited only to emergencies like the BP / Deepwater Horizon spill.

BP / Gulf Oil Spill – How Big?

The Macondo well is still tightly capped; the “static kill” effort to shut it down by pumping mud, then cement, into the well through the containment cap may begin as soon as tonight. This would be followed up by a “bottom kill” procedure, hopefully in August, that would use a relief well to pump more cement into the well at a point about 13,000′ below the seafloor.

So, folks are asking – just how big was this spill?

Big enough to – at one time or another – cover 68,000 square miles of Gulf waters with oil slick or sheen, based on our ongoing analysis of satellite images. But that’s just the part we could see at the surface. Lots of oil remained in the water column, beneath the surface, out of sight on the satellite images we’ve been able to acquire.

The Washington Post published a calculation on July 29 (article and useful graphic) that as much as 218 million gallons (5.2 million barrels) leaked out of the well over the duration of the spill from April 20 – July 15. That assumes the government team’s high-end leak rate estimate of 2.52 million gallons (60,000 barrels) per day for 87 days. Subtracting 33.6 million gallons (800,000 barrels) the Coast Guard and BP claim to have kept out of the water — by siphoning oil directly from the leaking well — yields a high-end spill estimate of 184.4 million gallons (4.4 million barrels).

SkyTruth’s estimate on May 1 that the well was gushing at a rate of at least 1.1 million gallons (26,500 barrels) per day turned out to be on the low end of the later scientific estimates made by the government-assembled Flow Rate Technical Group. Our conservative number generates a total flow of 96.8 million gallons (2.3 million barrels) from the leaking well over 87 days. Subtracting the 33.6 million gallons supposedly diverted from the leaking well – we have no way to confirm that number – yields a low-end total spill estimate of 63.2 million gallons (1.5 million barrels) directly into Gulf waters.

How does this compare with our previous sad benchmark, the Exxon Valdez spill in 1989? The official estimate — a number that, we should note, is disputed as being far too small — is that 11 million gallons were spilled when the Exxon Valdez supertanker ran aground in Prince William Sound, Alaska.

That would make the BP / Deepwater Horizon oil spill in the Gulf anywhere from 6 to 17 times as large as the official Exxon Valdez spill. But don’t lock these numbers in just yet: they may change as the investigation proceeds and new information is brought to light in coming months.

We’ll have to work hard, and stay on guard, to make sure this is one sorry new record that will never be broken.

Twitter Not Working (For Us, Anyway) – Oil Spill Updates

For some reason we’ve been unable to tweet for more than 24 hours now. We’re entering severe withdrawal, but we can still give you this update (in waaay more than 140 characters):

  • Michigan pipeline spill / Kalamazoo River. The EPA estimates this spill exceeded 1 million gallons, and has now traveled more than 35 miles downriver from the point of origin since the leak began on July 26. Michigan governor Granholm is urging more aggressive response to keep this spill from reaching Lake Michigan.
  • Louisiana blowout and spill / Barataria Bay – Bayou St. Denis. The Coast Guard is saying it will take at least 10-12 more days to plug the abandoned well that has been spouting a 100′ geyser of oil and gas out of water since it was hit by a barge on July 27.
  • Dalian, China pipeline explosion and spill. Greenpeace claims this spill is much larger than reported by the Chinese government – possibly 60 times bigger, based on revelations that Chinese workers purposefully dumped oil into the ocean so it wouldn’t feed the raging inferno and cause more destruction of storage facilities onshore. Greenpeace also claims a full oil storage tank capable of holding about 28 million gallons was destroyed during the fire, possibly releasing its contents into the water as well.
  • BP / Deepwater Horizon spill, Gulf of Mexico. The containment cap is holding, remains shut, and no new oil has leaked into the Gulf from the Macondo well since July 15. Although thick, “skimmable” oil slicks have reportedly become hard to find floating on the Gulf’s surface, questions remain about how much oil continues to linger beneath the surface and out of sight. Recent satellite images show what we assume is mostly thin sheen still present across a large area. The much-anticipated “static kill” procedure to pump drilling mud directly into the well through the cap is now planned for Tuesday, with the relief well in position to begin intercepting the Macondo well by August 11 or 12. Successful execution of the “bottom kill” procedure – pumping more mud, then cement, into the well via the relief well – could take an additional three weeks.

Today’s MODIS / Aqua satellite image of the Gulf seems to have good illumination conditions for showing oil slicks and sheen east of the Delta. We don’t see much indication of the widespread sheen that was present on the July 28 imagery, although a large part of that oily-looking area is south and slightly west of the Delta and obscured by clouds on today’s image. Stay tuned, this continues to be a very dynamic event.

BP / Gulf Oil Spill – Curb Your Enthusiasm, Part 2

Yesterday’s MODIS and RADARSAT images show something we didn’t expect: slicks and sheen spanning nearly 12,000 square miles. Based on other reports, and the recent trend on satellite images indicating steady dissipation of the surface oil slick, we are optimistically assuming that nearly all of this is very thin sheen.

Speculation: winds from Bonnie obliterated most of the thin sheen throughout the area; but since then, sheen has had time to “reassemble” into observable layers that noticeably affect the sunglint on MODIS images, and the backscatter on radar, but may not look like much to folks out in the Gulf on vessels or in low-flying aircraft. That’s our theory at this point. Chime in if you have other thoughts about what we’re seeing on these images:

MODIS/Aqua satellite image, July 28, 2010

The MODIS / Aqua satellite image above, taken at 2 pm Central time on July 28, shows oil slicks and sheen (encircled with orange line) that we think are likely attributable to the BP / Deepwater Horizon oil spill, spread out across 11,832 square miles (30,644 km2) in the Gulf of Mexico. We’ve marked the eastern edge of a persistent ocean-color anomaly with a dashed line; this anomaly may simply be related to the Mississippi River discharge, or could indicate an area where ocean chemistry has been affected by oil, dispersant, and/or dissolved methane from the spill and cleanup response. Three small slicks attributable to natural oil and gas seeps are also marked.

RADARSAT-2 satellite image (black-and-white) taken July 28, 2010. RSAT-2 data courtesy CSTARS.

We overlayed the RADARSAT-2 image (black-and-white) taken at 6:48 pm Central time on the MODIS/Aqua image taken earlier that same day. The large dark area on the radar image is probably oil slick and sheen from the BP oil spill: wind conditions throughout the area were ideal for slick and sheen detection on radar satellite imagery, ranging from 2 to 8 meters per second with minimal gusts. Weather satellite images taken at about the same time showed few clouds in the area and very low chance of any rain in the vicinity.