Saturday, April 26, 2014

More NDT

Laboratory Technology / Patriot Services

This piqued my interest lately: "Lab Owner Sentenced in Connection with Pollution Reports"
The lab company was faking brine shrimp lethality tests for a period of at least 4 years. Those tests (technically "bioassays") are important for determining that produced water discharges are within accepted limits. 

What is a brine shrimp lethality test?  It's a simple, widely-used bioassay.  At it's most basic: add water, add shrimp, add substance, see if substance kills shrimp.  Wikipedia.  It's so easy to perform, you could teach a rhesus monkey or a grad student to do it.  Here's a nice description from an NIH paper:

"The brine shrimp lethality bioassay is rapid (24 h), simple (e.g., no aseptic techniques are required), easily mastered, inexpensive, and requires small amounts of test material (2-20 mg or less).[12] The bioassay has a good correlation with cytotoxic activity in some human solid tumors and with pesticidal activity.[12,13] This test was proposed by Michael et al.[14] [NOTE: 1952 paper; these tests have been used for >60 years] and modified by others.[15,16] Since its introduction, this in vivo lethality test has been successively employed for providing a frontline screen that can be backed up by more specific and more sophisticated bioassays once the active compounds have been isolated."

A good quality lab-grade scale costs about $300.  One of the writeups up top mentions a 'broken scale' that wasn't fixed...  Not fixing a scale to save costs doesn't wash.  There's more to it than that.  

For a quick digression, I noticed the company next door to Laboratory Technology is Patriot Services. Patriot vends fluids for drilling, including solvents and loss circulation fluids. Remember how on Macondo, they had leftover fluids they shoved down the well as a spacer so they didn't have to pay for hazardous disposal? Also, Google Street view shows a number of chemical tote tanks at Patriot Services.  I looked at the MSDS sheets and most of what they vend is (according to the MSDS sheets) fairly benign.  

Looking on the Secretary of State website, Laboratory Technology and Patriot Services happen to be the the same company.  The agent for all of the filings over the years was Jim Donelon, the Insurance Commissioner for Louisiana.  E. Gerald Hebert, the owner, happens to be an active political donor (at least $57,000 between '98 and '12, exclusively to Republicans), including to Bobby Jindal.  Yet, Patriot Services also won $130k worth of Stimulus Act funds.  

A company named "Patriot Services" based in Kenner that sells proprietary drill fluids and is owned by a Jindal donor?  That's not the slightest bit suspicious.  

I'd also like to highlight the recent, infamous "coffee filter" conspiracy.  Hmm... 

The charge that she plead guilty to is a felony, yet she only received probation and a fine... Is this a deal where the Feds will work their way up the ladder?  Who were the customers of this lab (I've found at least one: Virgin Offshore USA, who had "the worst measured safety record in the Gulf")?  What did they think when every single sample sent in for years came back clean?  Where was LA DEQ?  There are many interesting questions that can be asked about this incident if someone wanted to start poking.  

Louisiana Chicken Boxing

Sunday, April 20, 2014

Courtwatch, BP Edition, Part 3

Four years ago, what we'd come to call the Macondo blowout / Deepwater Horizon / the BP Oil Spill occurred. The Atlantic / 4 Years Ago, Photo Essay
Previous Edition: Part 2
First off, a quick conviction update: A BP manager who was in charge of spill response has plead guilty to insider trading charges. His was an obvious crime. You've got to be pretty blatant to get the SEC to come after you. He sold all of his shares very early on after the well blew out and all of his family's shares. That's a little too flagrant for the SEC to give you a pass.
Probably the phase with the most variability is the "how much spilled" phase (NY Times: In BP Trial, the Amount of Oil Lost Is at Issue - Nice summary). Note that many of the fines are on a per barrel basis, so, if more barrels were spilled, bigger fines. I believe it's $1,100/barrel under regular negligence and, if gross negligence is found, it's over $4,000/barrel. By fudging their estimates, it's argued BP can best limit their total liability. Here's more on the sparing over the spillrate estimates.
Before I go into what I viewed in the courtroom, I'll point out this very important factoid: Actually, they kinda DID have a meter. During the last stages of the blowout, a capping stack with 2 orifices of known size were fitted to the top of the well. One of the two orifices was fitted with a pressure gauge that discharged directly to the GoM while the other was routed to the collection system of the Disco Enterprise & Q4000. There was a pretty good gauge of how quickly the well was flowing at that point in time (~50,000 barrels/day).
The person (Dr. Tom Hunter) that noticed the lineup and made the initial calculations was actually a nuclear weapons engineer. He had a PhD from Wisconsin and held roughly the same position as Dr. Oppenheimer had. He worked for Sandia National Labs and was a part of a taskforce led by DOE to fix the blowout and monitor BP's actions. I actually got to sit in the courtroom for some of his testimony. He retired about midway through the blowout and decided to stick around to help make things right, despite being paid by nobody to do so. His testimony was pretty amazing. If a lawyer could sit down and dream up the most reputable witness possible, Dr. Hunter would be high on that list. He definitely spoke like an engineer (he wasn't very polished and kinda stuttered), but his command of the numbers was firm. His testimony cemented the government's position pretty well early on in the spillrate phase of the trial. BP didn't spend a lot of time cross examining him and was probably just glad he was off the stand. I would have liked to see BP try impugning his integrity, after all, Dr. Hunter was just in charge of making sure US Nuclear Weapons go boom. The best testimate to Dr. Hunter's integrity is that his calculation, which was made with not much more than a 4-function calculator in the field while the blowout was going on, was never questioned by anyone, even scientists and engineers on BP's dime.
Here's the final government report on the spillrate:
Federal Government Report on Oil Spill [PDF]
For reference: ~4.9 million barrels total
Note that the government narrowed in on a single number which was then subjected to peer review and was published far in advance of the trial. For a ballpark figure, it amounted to an average of ~50,000 barrels/day.
In the leadup to the spillrate portion of the trial, BP played a little dance with the numbers. They constantly complained that the government's figure was too high, but not saying by how much nor would they expose their own reports until just before the spillrate portion of the trial. Off the top of my head, it wasn't until about 3 weeks prior that they posted their number on their website.
BP Report on how much oil spilled
For reference: ~3.26 million barrels total
Note that BP actually had about 3 different numbers, all of which coalesced around ~40,000 barrels/day. Both the Feds and BP agreed on the total oil collected by the capping stack before the well was finally shut off.
The initial shock was, 'wow, BP's number isn't much different from the Feds.' Everyone expected a super-lowball estimate from BP, but actually (in large part because of Dr. Hunter) they were pretty hemmed in by that final flowrate measurement, which everyone agreed was about as accurate as possible under the circumstances. The biggest difference between is the Feds assumed a constant or decreasing flowrate; BP's report argues that the flow rate kept increasing over timer due to erosion. Since the endpoint is fixed, they were just arguing over the slope of the curve (pointing up or pointing down). The Feds claimed that reservoir depletion was the governing factor. BP argued that erosion of the partially-closed BOP rams was what governed the flowrate.
As I mentioned, I saw some of the Federal witnesses (the Feds went first). I also got to see a couple of the BP flowrate estimate witnesses. The later was using an OLGA model. He was pretty reputable. He had (as memory serves) the highest of the BP estimates.
The first BP witness, however, was a haughty British geology professor out of Imperial College of London. He had the lowest of the BP estimates and depended on reservoir modelling for his results. He used equations of his own derivation based upon first principles. I got to witness some of the cross examination which (I think) was done by the same young prosecutor that led Mix's prosecution. He tore into the Brit professor, going into how much money BP had donated to Imperial College, how he used to work for BP, how his flowrate estimate was only lately published and wasn't peer reviewed. The professor, who didn't handle the pressure well, retorted, 'well, it was posted on the internet and I've received some really nice comments from people.' Not exactly as impressive as a peer-review.
What will Barbier do? The levels of uncertainty on this measurement are far beyond what scientists are usually comfortable with. There's no question I think that the Federal estimate is more reliable, but if Barbier would just take the mean of the two estimates, I wouldn't have a big problem with that. How much oil spilled?
Billions of dollars are on the line (a $16 Billion dollar question). Many of those dollars are reserved for coastal restoration in Louisiana. What choices Barbier makes will be worth paying attention to for a number of reasons.
NOTE: Some minor edits after publishing.
UPDATE: The dates for the final phase of the main BP trial were just announced today:
Richard Thompson (@rthompsonMSY) tweeted -
"Mark your calendars: Phase 3 of #BP oil spill trial, the "penalty phase," to begin 1/20/15 and end 2/5, per order issued this morning."

More Patches

Patches Update: Keeping a charge up Edition

Patches makes a friend who is even rustier

Ever since I bought Patches, it's had a slow drain on the battery. The wiring is such a rat's nest, I can't even begin to troubleshoot it, so instead I just resort to trickle charging and jumping Patches on a not irregular basis.

I've had an EZ Wiring Harness for a while, and that's the permanent fix, but that'll take a while (and takes Patches out of commission in the interim).

For the meantime, I've gotten a temporary fix in place:

A terminal post disconnect switch. Note the trickle charging in progress as well.

Also, with the help of New Orleans Classic Motorworks, Patches no longer leaks oil! Yay! The oil pressure sense tube wasn't screwed in correctly at the back of the engine (behind the distributor). It's a bitch to get to and it also sprayed oil all over the back of the engine and was a total mess. After re-running the line (and a new oil leak at the gauge, which was quickly fixed), no more leak.

Thursday, April 17, 2014

Neil DeGrasse Tyson at Tulane

We arrived early for Neil DeGrasse Tyson at Tulane, but barely made the overflow room. Summary of the talk. Excellent summary. And then the comments... Sigh ... Louisiana needs some stupid knocked out of it.

I've been poking fun at coworkers who went to LSU saying I don't think that they could have managed hosting NDT ("Dr. Tyson, what's your opinion of chicken boxing?"), while Tulane could. Tulane's Campus programming even ate his speaking fee and opened it up for the public (yes, there was a hell of a line, but kudos to Tulane).

I also want to relay a little story. Not only is Neil DeGrasse Tyson a brilliant astrophysicist and the coolest scientist alive, but he's also a tremendous human being.

We bumped into a friend of mine at the overflow room. He had his wife and ~6 year old daughter there with him. As little ones are wont to do, the daughter had to go pee 5 minutes before the event started. The wife & girl hurry off. While they are gone, Dr. Tyson darted into the overflow room, waved at everyone, said Hi, then started running back across McAllister Drive for McAllister auditorium. As he was running through the student union, he saw my friend's daughter. Dr. Tyson stopped in his tracks, turned around, ran back to the little girl, high-five'd her, turned back around, and then hustled back to start the talk. He talked for an hour and a half then took questions.

Update: videos