Forensic Analysis of BP’s Bathymetric Chart
Forensic Analysis of BPâs Bathymetric Chart
by BK Lim, Geohazards Specialist
(2 Sept 2010, email@example.com).
(updated 7 Sept 2010, correction to true slopes due to misread contour interval)
For a disaster of this magnitude, many things must have gone wrong from the very beginning. The long chain of human errors leading to the disaster could not have been purely accidental or random in nature. There must have been willful negligence, ignorance or misinterpretation to capitalize on certain opportunities or to make good on some bad situations along the way.
Like many, I followed the intense discussions at The Oildrum.com (TOD) to get some fair technical coverage of the BPâs oil spill disaster. As the disaster wore on, I started to wonder why industry experts like Art Berman, Rockman and many supporting actors (Rocdoc, PinkFud, Quaking, CraigWcoop & others) were so defensive of BPâs lies and zealously stamping out any independent bloggersâ views. If they had been truly professional and interested in seeking and disseminating the truth of the disaster, why were they not discussing incriminating issues that BP seems to be avoiding? Granted that nuking the gushing well was not really a good idea, but why should they be bitterly thrashing Matt Simmonsâ apparently valid assertions as well.
Being a geohazards specialist, I started my own investigation with the bathymetry as I had always done. Being the lowest denominator (in terms of data ranking of importance and sophistication) any evidence of willful negligence should be readily apparent. To my surprise, I found many issues with the bathymetry which nobody seems to be discussing at all. So in the first draft of my posting; Would a better bathymetry chart have made the difference?, I asked the question âIs there something wrong with BPâs Macondo bathymetry?â
There had been some initial troublesome scaling errors while trying to superimpose BPâs bathymetry onto the satellite image from google. I could not find the reference points and was unfamiliar with a prospect using US imperial system. Instead of correcting it (since it was inconsequential to the development of the geological model) the scaling error was left uncorrected in the initial posting. It looked like an opportunity to use it as a âcontrolâ to gauge the readersâ responses on the erroneous Macondo bathymetry.
The five areas of fundamental problems with BPâs Macondo bathymetry are:
1. Data acquisition fraud: BPâs Macondo âsmoothenedâ bathymetry did not appear to be consistent with MBES (multibeam echo sounder) images scanned at close range to the seabed. The bathymetry appeared more likely to have been compiled from widely spaced SBES (single beam echo sounder) data along surveyed lines. MBES mounted on AUV is designed to optimally scan the seabed with multiple ultrasonic beams (>100 to 250 beams) from an altitude of 150-200 ft above the seabed, while the conventional SBES utilizes a single ultrasonic beam from the sea level (5000 ft above). Needless to say, data acquired using the AUV-mounted MBES would yield a much higher resolution image of the seafloor as compared to one using a SBES. Three seafloor images using comparable 3m-bin size MBES data were included in my earlier posting (Would a better bathymetry chart have made the difference?) for comparison.
From my experience, passing the backup SBES data as MBES (a form of cheating) is widespread in the industry. Imagine a multimillion AUV-MBES system failing to function in the middle of a survey out in the open sea. Would the survey contractor stop work, return to port and resume survey after the system had been repaired? Sometimes it is not the question of cost. Whether time would permit is another important issue. In most cases, the survey contractor would rather continue the survey using the backup SBES system than to lose a few million dollars. With survey acquisition costing between 50,000-75,000USD/day, the compelling choice is clear. Whether the BP company man onboard the vessel knew or was happily sleeping in his cabin is open to question. Of course, 90% of the survey contractors would never openly declare that their AUV mounted MBES was non-operational 4850ft below the sea surface. Their favorite fallback argument has always been âWhat is the fuss as long as the bathymetric chart can be produced?â When the bottom line is at stake, quality and resolution fly out the window. MBES can cost more than 10 times the cost of SBES data acquisition.
2. Data compilation fraud: At the post survey data compilation stage, few survey companies would openly admit their data acquisition shortcomings. So although the bathymetry for the Macondo prospect was compiled mainly from SBES data, the legend in the chart still states âMultibeam Processing Sequenceâ. The graduated colour scheme normally used in MBES images (unusual for SBES compilation), is another tell-tale sign of willful intent. There are other evidences but these should suffice for the purpose of this article.
3. Willful Negligence and Ignorance: For the benefit of the layman, horizontal (spatial) resolution refers to the smallest measurable quantity or interval laterally. As the ultrasonic beam gets further away from the source (echo sounder), the reflection circle (cone) gets wider. Thus a SBES system using 30kHz frequency deployed at sea level would not be able to resolve or see better than 4.5 m or 14.5 ft (spatially) at depths of 5000 ft below sea level. The bin-size (9.84 ft) and contour interval (5 ft) cannot be smaller than the resolution limit of 15 ft. For comparison, a 30kHz SBES source deployed 150 ft above the seafloor would have a spatial resolution limit of only 2.5ft.
Mathematically, the minimum contour interval should be at least 15 ft but for practical mapping purposes, it should be twice the resolution limit or 30 ft and not 5 ft as BP had used. This means any point on the map has an uncertainty circle of at least 15ft radius. Contouring at 5 ft interval (1/3 the resolution limit) would have unduly distorted and biased the compiled bathymetry; the survey lines being 660 ft apart or 132 times the contour interval.
In contrast, a 3m-bin size means there is an average depth point (calculated from several angled-beams) at every 3 m cell. For a SBES source to generate the same data density, the Macondo bathymetric survey line-spacing has to be 3m or 66 times the number of survey lines shown on the chart. But surveying the lines so closely does not make any sense with a spatial resolution of 15 ft.
Would you trust BPâs pledge âto make it rightâ in the aftermath of the disaster when they could not be bothered with doing it right even with the fundamentals? The cost of a proper bathymetry survey is probably 1/100th the cost budgeted for well A. Is this a classic case of penny wise, pound foolish?
The lack of resolution will also explain the smoothness of the seafloor which is inconsistent with true slopes exceeding 3Âº. Whether this was willfully overlooked is open to question. If the chart is to be used, the resolution limit and accuracy should have been clearly stated and cautioned. In BPâs permit application, the depth for Well A and Well B was stated as 4992 ft; implying an accuracy of 1 ft. This is willful negligence. Accuracy cannot be better than resolution. What is the purpose of implying 1 ft accuracy when a more honest representation as 4990 Â±10 ft would suffice. Why was there a need to impress (or mislead)?
If AUV-mounted MBES data had been used as specified instead of the backup SBES sounding from the sea surface, the bathymetry would have been more accurate and better resolved to show the irregularities consistent with the true steep slopes (>3.7Âº to 6.6Âº) of a major escarpment. With a more accurate and detailed bathymetry, BP should have seen the minute tell-tale signs of the hazardous conditions beneath the worst possible parts of the escarpment to drill. If the well had been drilled from a safer seabed location, perhaps the Gulf Oil Spill disaster might not have happened?
4. Willful Misinterpretation: As evident in figures 7 and 8 of Satellite Image Comparison, the L-shaped escarpment which covers almost 2/3 of the Macondo site, stretches 12 km to the NW (~0.6 km wide) and 5.5km to the north, with the width of the north-eastern flank varying from 1.3 to 1.6 km. Although the total height of the escarpment cannot be measured from the satellite images, the lower slope itself is already 250ft. Compare this observation with BPâs assessment:
BPâs Shallow Hazards Assessment:
The only seafloor feature identified on the exploration 3D seismic data within the vicinity is a low-relief escarpment approximately 1,000 ft to the south of âAâ location which is the seafloor expression of a deeplyâburied scarp associated with mass-wasting.
The only seafloor feature identified on the exploration 3D seismic data within the vicinity is a low-relief escarpment approximately 950 ft to the south of âBâ location which is the seafloor expression of a deeplyâburied scarp associated with mass-wasting.
BPâs assessment was obviously very far from the truth. The escarpment is definitely not the only seafloor feature. Even from the satellite images and the âsmoothenedâ bathymetry, there are obvious features such as steep to gentle slopes, almost flat seafloor at the canyon bottom, topographical irregularities etc. No one can deny that both Wells A and B are on the mid-slope of a massive escarpment rather than 950-1000 ft north of it.
Yet Fintan Dunne and Art Berman (TOD), both respectable experts in their own fields, can still assert that well A is located within a valley. Preconceived minds do work wonders.
This discrepancy can only mean that BPâs geohazards assessment had misinterpreted the southern foothill of the Massive Escarpment as the âescarpmentâ itself. See BPâs misinterpreted escarpment in figure 115-1. How can a massive escarpment be interpreted as an âedgeâ in the middle of a slope with no significant change in gradient? Was the steep irregular topography willfully downplayed to deceive the regulatory body (MMS) into approving the exploration application?
In my 30 years of geohazards work, it is totally illogical for a blowout to be so disastrous and yet the site can be described as âgentle and featurelessâ as implied in the hazards assessment. Did BP totally ignore the original unfavourable geohazards assessment? This question can only be answered after an independent review of the 1998 and 2003 geohazards reports. Why would BP America Inc carry out its own internal mapping in 2008 and 2009 using exploration 3D seismic data when there were already existing geohazards reports? Exploration 3D seismic data meant to map more than 25,000 ft below sea level, do not have the necessary resolution to detect and resolve shallow geological hazards as many had experienced in the past of such âdisastrous experimentsâ. If BPâs internal mapping using 3D exploration data took precedence over the previous shallow geohazards reports in any way, then it would be difficult for BP to wriggle their way out of willful negligence?
All the previous shallow geohazards assessment and site survey reports should be reviewed and examined for evidence of willful misinterpretation and lack of due diligence in the shallow geohazards assessment.
5. Willful Misrepresentation: Figure 115-1 shows the seabed profile along the line XY drawn diagonally across the bathymetric chart. BP stated the slope dips 3Âº SE. Even with the current terrain âflattenedâ by resolution limitation the average true slope is already 3.7Âº. In slope analysis, it is the steepest slopes that matter, not the gentlest. It is also erroneous to quote a generalized slope since the slopes of the escarpment is neither uniform nor dip in the same direction. True slopes currently measured from the erroneously smoothened contour vary from 1.7Âº to 6.6Âº. This is certainly not uniform. In reality, the true localised slopes should be a few degrees higher than presently measured.
Why did BP state in its report that the seabed slope is ~3Âº and the escarpment is low relief? Even 3Âº slopes are considered to be steep, as most well locations are located at seabed with 3Âº should have been investigated for possible shallow hazards, especially when the location is located right at the convex face of the escarpment. Any escarpment or raised landform exceeding 250ft (75m) in height cannot be described as âlow reliefâ? BPâs choice of words and terms were definitely misleading? The cursory hazards assessment given in only one paragraph (less than 100 words) certainly do no justice; even if the severe slope is the only consideration.
Missing the Forest for the trees
If there had been no comments at all from TODâs industry experts on my bathy posting dated 24 July 2010, I would not have been so convinced of BPâs deceits and willful mass deception to cover up the long chain of human errors leading to the disaster. Even though there had been visits from BPâs goons to my column and in particular the Bathy posting, they did not want to bring more attention to my blog by commenting.
On 29 July2010-10:30am, Robert Rapierâs posted âA Critical Examination of Matt Simmonsâ Claims on the Deepwater Spillâ. It must have been a tremendous group effort involving multiple Oil Drum staff members particularly Joules Burn, Art Berman, Euan Mearns and Robert Rapier, to compile all those satellite photos, statistics and causes of methane emission (including belching cattle) just to put down one man.
Just when they thought they had successfully âsnuffed outâ the last flame of argument against BP (after an exhaustive day hammering down Matt Simmons), someone posted my article, Why Is BPâs Macondo Blowout So Disastrous & Beyond Patch-Up? for discussion the next day. My diagrams and geological model must have infuriated those BPâs goons at TOD. Not only were my diagrams more beautiful than theirs, many impartial bloggers seemed to think my geohazards assessment made a lot more sense. It must have been most frustrating for those BPâs goons to have another sprout of doubts springing to life, so soon after the last one was snuffed out.
It could not have been coincidental that the debasing comments on the inconsequential scaling error by GeoNola, Klurker and Lurking came one day after my diagrammatic illustration of BPâs Macondo Blowout had attracted a 6,000% jump in website visits (first day) to my column. my diagrammatic illustration of BPâs Macondo Blowout had attracted a 6,000% jump in website visits (first day) to my column. Their complete silence on the fundamental issues was deafening. All except Fintan Dunne are oilman experts and judging from their blog comments, are fiercely defensive of BP.
Thereâs an old saying; âYou can see the ant over the river but not the elephant on the same bankâ.
I picked BPâs bathymetry to illustrate what many had long feared. It is an open record of BPâs list of willful negligence which everyone can understand. None of BPâs goons, who had been so vocal on so many issues, have remained âloudlyâ silent on issues incriminating BP. This proved beyond reasonable doubt who they had been working for. I will end this posting by quoting the latest news of BPâs tripled record breaking spending on advertisement.
BP tripled Ad spending after spill – NYtimes
It will come as little surprise to newspaper readers and television watchers, but BP significantly increased its spending on advertising after the April 20 Deepwater Horizon explosion and oil spill. BP spent $93.4 million on newspaper, magazine, television and Internet advertising in the three months after the disaster, three times what it spent in the comparable period in 2009 the company reported to Congress.