Fix the pumps

Thursday, May 31, 2007

How about briefing the citizens and taxpayers?

[Updated July 11, 2007. See below]
[Updated June 9, 2007. See below]
[Updated June 1, 2007. See below.]

From the May 31, 2007 Times-Picayune, in an article regarding General Van Antwerp's visit to New Orleans:

"...The general will be briefed on the problems the corps experienced installing pumps at three temporary gates built at the ends of the 17th Street, Orleans and London avenue canals."

I assume this is in conjunction with the internal review of the pumps, which has been proceeding for nine months. Do we seriously think the general is going to receive the same patronizing BS that the public gets, like this May 24, 2007 gem from Colonel Bedey (talking about a pump going down during Orleans Avenue pump testing)?

"I equate it to an alternator going out on a car. You don't replace the car, you replace the alternator"

Thanks for fully explaining the problem to us. It's really inspiring that you think that none of the citizens are intelligent enough to absorb a complete explanation. You couldn't put out a press release the next day, or even in the week since then? Do we even know if the pump has been fixed? Do we know what the problem was? Do we know if the same problem can be found on the other 39 pumps? "No" to all four questions. It's typical of Colonel Bedey and his band of Merry Mendacity-Makers to not level with the public. Happy talk and collegial back-slapping has grown very tiresome. We deserve unvarnished, unspun facts. Apparently, Van Antwerp is far more likely to receive them than the people who actually need them. This is not openness.

Update, 6/9/07:
By the way, the Times-Picayune found out it wasn't a matter of a simple "alternator." Buried at the very end of their article about the release of the Corps internal investigation was this howler:

"During a news conference on the report Friday, Bedey also confirmed that one of the new pumps was replaced May 17 after a chunk of concrete was sucked into it, causing severe damage. He said the replacement took two days."

A chunk of concrete!

This also brings up the interesting question: where did they get a 60" pump in just two days? They didn't buy any spares.
[end update]

The citizens of New Orleans, who have to live with this stuff, should get the exact same briefing that a general who has been on the job for less than two weeks - and who does not have to live behind those gates - should get. Frankly, New Orleans should get that briefing before the general, even if it might embarrass the general. And local and national reporters should be crawling all over every office of the Corps (New Orleans, Vicksburg, or HQ in Washington) that might have access to the briefing and the internal review report in order to get them to the public. We deserve it.

General Van Antwerp's predecessor, General Strock, told Senator Vitter the Corps internal panel would get their work done by the end of May. Well, today's May 31. If there are recommendations and action items in there, wouldn't it be important to the citizens to know them, so they can make informed decisions about the risk they face this hurricane season? That season starts tomorrow.

By the way, it is very likely that there are recommendations and action items in the Corps internal report. The Corps doesn't investigate for the sake of investigating. They are a militarily-goverened organization, and as such, are action-oriented (at least in theory). They would not spend nine months on an investigation, and then say "Everything's fine!" At least I hope not. My instinct is that there are things in the internal report that show there is still work to be done on the MWI pumps, despite all the puffery from Col. Bedey on the local news.

Update, 6/1/07:
At the very end of the Times-Pic's coverage of Van Antwerp's 17th Street canal press conference is a tiny mention of the internal report:

"Van Antwerp also said an independent report on the problems experienced by the corps in installing pumps at the ends of the drainage canals should be released next week. The report was requested by [Corps Mississippi Valley Division Commander Brigadier General Robert] Crear after inspectors raised questions about the ability of the pumps to operate."

Separately, I have heard that Crear is writing/editing his preface to the report, and that is part of the holdup.

I did notice that Van Antwerp came out smelling like roses (mostly) in the TV coverage yesterday and in the print coverage today, so I guess that was "mission accomplished" for the Corps' Public Affairs operation. Though I must give the T-P props for calling him to task for a bit of cluelessness:

"Van Antwerp admitted he knew nothing about the long-delayed Hurricane Protection Decision Chronology being conducted under the auspices of the U.S. Defense Department.

The study was designed 'to enable the corps and the nation to fully understand the long history of federal, state and local decisions that led to the design and construction of the New Orleans-area flood and storm damage reduction system,' Maj. Gen. Don Riley, corps director of civil works, said when the study was announced in June 2006 as a companion to the nine-volume forensic report on the levee failures during Hurricane Katrina.

Corps officials and members of the independent team preparing the report have refused to say when it will be released to the public. The lack of such a report on the breaches -- detailing policy and political failures that accompanied engineering and scientific failures -- prompted criticism last year from the American Society of Civil Engineers."

In order to promote accountability later on (in case stuff goes to hell), here's a wrapup of Van Antwerp's comments during his visit:

Times Picayune: June 1
AP: New corps chief says he's personally committed to rebuilding New Orleans
WWL-TV (video only, report from 5 PM news): New head of Corps of Engineers in town
ABC-26: New Corps chief tours area
WDSU-6 (video only, report from 5 PM news): Corps Chief Promises to Tell Truth, Good or Bad
Baton Rouge Advocate: Storm risk report to be very detailed

[Update, July 11, 2007]
The Corps released the Katrina decision chronology today. The Times-Picayune did a big article that carried the Corps water quite adequately. Despite actually naming names within the article, the headline and the lede still say that blame can't be assigned to individuals within the New Orleans District. This is crap. Real people signed those drawings and wrote those reports and advocated those positions that led to Katrina. Some of those people are still working in the New Orleans District. So much for accountability.

Hopefully, all of that will come out in the civil suits.

Tuesday, May 29, 2007

GAO report released

Updated 7/2/07. See below.

The GAO has released their report on procurement of the 34 floodgate pumps from MWI. It is available here (or here). The official title is "U.S. Army Corps of Engineers' Procurement of Pumping Systems for the New Orleans Drainage Canals" and the GAO's number is GAO-07-908R.

This is not the final version of the report, but rather the briefing slides presented to Senator Landrieu's office on May 17, 2007. A so-called "blue cover" (i.e. full-blown) report will eventually be issued.

I'm reserving judgement on the GAO's work until the results of the Corps' independent review are (hopefully) made public. In his May 2, 2007 letter to Senator Vitter, now-retired Lieutenant General Strock (Corps Commander) mentioned that the independent review was due to Brigadier General Crear (head of the Mississippi Valley Division, which is one layer above the New Orleans District) by the end of May, which is this Thursday.

You can read my thoughts regarding GAO's work in my July 2nd post here.

I'm a weenie

After a fiasco in the comments to this post a couple of months ago, I got a lot more selective in my moderation of comments. Some people were writing some really scurrilous dirt, and I cut it off. I realized that I am under no obligation to give individuals a forum to sling personal insults or to make vague accusations with nothing to back them up. Also, I am not interested in injecting politics into what I write. In fact, I'm not interested in injecting anything personal into what I write. Finally, I'm interested in comments that are written well, without lots of SCREAMING or exclamation points.

Despite all this, and the fact that "Comment moderation has been enabled" is written on every comment screen attached to every post (and has been since I started blogging last September), some people persist in the belief they have an absolute right to see everything they write published on someone else's website. That's not how the real world works.

I should make clear that I don't discriminate against anyone in particular when moderating comments. In reality, unless a commenter identifies themself in the body of the comment, I really have no idea who they are. Most people do not attach their names to their comments in the body of the text; a few who have webpages fill in the webpage field. That's fine. There's no obligation to identify oneself when criticizing another person, though I believe it is the polite thing to do if one is interested in dialogue. From a practical standpoint, it also helps to distinguish oneself from all the other comments which have "Anonymous" as their author.

To those of you whose comments I've deleted, I'm sorry. If you feel it impugns my credibility not to publish everything that comes across my desk, you are entitled to that view. I read every comment and judge it on its own merits before I pass it through to the great, wide world.

However, as I said in the comments to that March post, this is my sandbox. Like every other blogger, I reserve the right to publish whatever I like. You are free to agree or disagree. I view the comments fields as more of a "Letter to the editor" forum (call me a Luddite for still reading newspapers). Newspaper editors do not publish every letter they receive, nor do I publish every comment I receive. That's what comment moderation is about, and it shouldn't really be a surprise. I've had moderation turned on from my very first post. There are more than enough other forums on the internet where people can complain about anything under the sun with impunity. Heck, starting a blog literally takes five minutes.

If anyone has a deep problem with what I'm writing, and wants to really engage in a rational discourse, they're free to email me. My address is in my profile, which is linked at the upper right of this page. Due to a mistake on my part, for a very long time, my address wasn't there, but I've since fixed that oversight.

In an exception to my policy of not posting stuff that's personally insulting, loaded with politics, and not at all on topic, below you will find the latest comment I received. According to the timestamp, it came at 8:55 this morning. It's attached to my post about the flowmeters. It is typical of stuff I've avoided posting, though I'm sure someone will have archived all their comments and will post them somewhere to show what a horrible person I am. Judge for yourself whether this is on topic and germane to a rational discussion. There was no name attached.

You are such a little weenie. I just came back to your blog and you deleted all of my critical comments.

You don't really want to post the peer-reviewed REALITY here do you?

I can't stand the Bush mobsters, but I REALLY despise dishonest engineers like you that post their own opinion, declare it to be absolutely gospel science, and then remove/dimiss all respondent commentary!

YOU are as bad as the BUSH PEOPLE!!

Tuesday, May 22, 2007


UPDATE, 6/1/07: Make sure you take a look at the comments to this post. There's more information there.

Each pump at the floodgates has a flowmeter on it which measures the flowrate of water through the pump units. The meters are ultrasonic units manufactured by Endress+Hauser. Specifically, they are model Prosonic 93W meters. What that means is that the transmitter (the box with all the electronic guts and the display) is a Prosonic 93. The "W" refers to the type of sensor. Endress+Hauser's flowmeter page is here. At that link, the specific page for the Prosonic 93W can be found by clicking on the box corresponding to "Ultrasonic Flowmeters" and "Conductive Liquids". Then, along the left hand side, you will find a menu entry for "Prosonic Flow 90W, 93W." Here's a picture taken from the Prosonic 93W page:

The way these things work is pretty simple. One of the sensors (the two little cylindrical things on the left in the picture, and which are about the size of a small can of tomato paste) is clamped to the outside of the pipe. It sends an ultrasonic waveform out through the pipe and the water flowing through the pipe. That signal is picked up by the other sensor, clamped to the outside of the pipe 180 degrees opposite to the first sensor. The other sensor sends the signal to the transmitter box (the blue and white box in the picture above, which is actually a model 90 transmitter - the model 93 has a bigger display and more plugs on the bottom) for processing. Depending upon how the waveform has changed as it has traveled through the water, an average velocity of the water can be calculated. If one wants a more accurate measurement, a second set of sensors can be added and the two signals can be averaged. The floodgates installations use just such a two-channel arrangement. After averaging, the transmitter can then send the flow data to a controls system for display elsewhere. The results are also displayed locally on the transmitter's screen.

It only makes sense that one would want as steady a flow as possible for measurement. Turbulence would scramble the signal, increasing the error in measurement. For this reason, the manufacturer recommends installation of the sensors a specified distance away from anything that could disrupt flow, i.e. introduce turbulence. Such things could be elbows, pumps, valves or pretty much anything other than straight pipe.

The distance is measured in multiples of diameters of the pipe. In the Prosonic 93 installation literature, which can be downloaded from this link, the specified distance downstream from a pump is called out as 20 times the pipe diameter. Here's the applicable picture from page 30 of the manufacturer's manual (which I have annotated in red):

The diameter of the pipe for each individual pump is 60". Those 60" lines - in groups of two or three - flow into 108" (9 foot) pipes that go straight out to the lake.

In the case of the flowmeters on each of the pumps, the sensors are located approximately 1 (one) diameter downstream from the pump. They are mounted pretty much right on top of the pump. Below are a few pictures to show the location of the sensors and transmitters.

Here's a picture of a removed discharge elbow at the Orleans Avenue site. These elbows fasten to the tops of the pumps once the pumps are in their frameworks out in the canal. I've annotated a few details on it. Inside the white box is the location for the sensors.

Below is an enlargement of the area inside the white box above. I've noted the locations for the sensors. As you can see from the photo above and the enlargement, they are only about four or five feet above the flange that bolts to the pump.

In the picture below I've noted the locations of the Prosonic 93 transmitters on the pump platforms.

Below is an enlargement of the transmitter on the left.

The picture below shows the Orleans Ave pumps on the east bank. I've highlighted the area which is enlarged in the next photo.

The picture below shows the highlighted area enlarged from above. As you can see, the flow sensor is just above the discharge of the pump.

In the photo below (an enlargement from another photo) you can see the orange straps on the adjacent pump which hold the sensors in place.

The photo below is taken from a short version of May 25, 2007 AP story about the testing of the Orleans Avenue pumps the previous day. The short version appeared in the Miami Herald, and had three pictures attached to it. The full version is here, sans photos.

On the picture, I've highlighted the bands that hold the sensors with red arrows. Inside the white box, one can see one of the sensors placed at exactly the location it was before the refitting of the pump units with hard pipe and new hydraulic motors.

And in the picture below, a screen capture from a May 15, 2007 WWL-TV report about pump testing at Orleans, you can see that the flow meters are definitely in use. The gentleman on the right appears to be a contractor from Prime Controls, according to the "PRIME" sticker on his hardhat. Prime is the contractor responsible for the entire SCADA system. I'm not sure what agency or company the gentleman on the left is from. But as you can see, they're both checking the flowmeter.

There's two problems related to the flowmeters here, one that may be solvable, and one that needs to have questions asked about it.

The first problem is this... in this installation, the flowmeters are pretty much in the worst possible spot to measure flow because of the highly turbulent nature of the flow immediately downstream of the pump discharges. As such, there is a large error being introduced into the flow measurement. In a perfect installation, the Prosonic 93 with W sensors has about a 0.5% error, which is pretty good. In this installation, which is far from perfect, the error could be significant. I can only guess, but I'm sure the Corps guys running the floodgates project know (at least I hope they do).

Is there a way to solve this? Maybe. There aren't many good spots on the individual pump discharges on which to place these sensors, due to the numerous sources of turbulence. For example, at 17th Street, all of the pumps shoot the water up into elbows, which are themselves bolted to elbows, which are welded to the 9 foot pipes. Thus, the individual discharges for each pump are nothing but sources of turbulence. It would be the worst place in the site to put flowmeters, but that's where they are. Here's a picture of 17th Street from last fall, showing the pump discharges:

Given a choice - and they are admittedly few - the best place to look for a location for the sensors would be on the straightest piece of pipe possible, as far away from sources of turbulence like the pump discharges or the tees where the 5 foot individual pump lines drop into the 9 foot pipes. The best place, given those parameters, would be on the 9 foot pipes on the lake side of the gates. According to the manufacturer's literature, the "W" type sensors can be placed on a 9 foot pipe and still work just as well. One would not have accurate flowmeter readings from individual pumps, but the existing flowmeters could still be used strictly for indication of flow ("It's flowing, but we don't know if the rate on the display is actually true."). However, one would have a better feel for the overall flow for the set of two or three pumps going into each of those nine foot pipes. Here's a picture of the lakeside location at 17th Street, taken from NPR's May 18th article about what's going on with the pumps:

Here's an overhead shot of 17th Street taken from a May 4, 2007 WWL-TV report, which shows the substantial length of straight pipe available:

The choice at Orleans isn't so great, since the straight section of the 9 foot pipes aren't nearly as long as at 17th Street (picture also taken from the May 18, 2007 NPR article):

As you can see, there are no sensors currently strapped to these locations.

London is about the same as Orleans, since it was designed by the same firm - URS (screen capture from April 30, 2007 WWL-TV report, sorry about the quality):

A closer view of the 9 foot pipes can be found in this USACE photo, previously shown in my October 21, 2006 post:

The second problem is more logistical in nature, but no less severe. One might think that the measurement of the flow is merely an academic matter. It's nice to know, but don't we just want to turn the pumps to "11" and let 'em rip? Well normally, yes. But remember that there's not enough pumps at the lakefront to match the what can be put into the canals on 17th Street and London Avenue. In addition, the Safe Water Levels must not be exceeded (By the way, that's going to be real tricky on London Avenue, considering that the pumps were specified to run at around 4 feet above mean lake level, which is also the safe water level on the canal. Modifications to the discharges of the 9 foot pipes have supposedly allowed the pumps to run at a lower height - perhaps 2 or 3 feet, but who knows if that's actually true?).

When one takes both these circumstances - too few floodgate pumps and the Safe Water Levels - into account, it becomes clear that there is a delicate balancing act to be peformed between the New Orleans S&WB pump operators in stations 6 (17th St), 7 (Orleans), 3 and 4 (London Avenue) and the Corps pump operators at each of the gate structures. The way that that balancing act is performed is by relying - in part - upon the readings from the flowmeters, which are piped through the SCADA system to the computer terminals in each pump station (except pump station 4 on London - I assume messages to the operators in that station will be relayed via the folks in station 3).

The Floodgates Operating Manual recognizes this. Here's the section titled "Step 9 (T - 6 hours)" (sorry, but the Corps isn't big on paragraph breaks. This is exactly as it appears in the manual):
"Conditions will rapidly deteriorate in the final six hours before landfall. Activities at the structure should be completed and all team members should be in safe and secure positions. The H&H Team will continue to monitor the storm and the predictions. The Canal Captains will need to closely monitor the canal levels since rainfall will likely begin to increase, leading to increased pumping at the canal pump stations. Through the working relationships established in the Water Control Plan, the Canal Captains will primarily be observers at the Local Partner pump stations but if the canal level rises to one foot below the Safe Canal Levels they will notify the pump operators and continue to communicate with the operators to verify they are adjusting pumping accordingly. If canal levels rise to the Safe Canal Levels then the Canal Captains will notify the pump operators and continue to communicate with them to ensure they adjust pumping to not exceed the Safe Canal Levels. The Canal Captains will remain in contact with the Operations Chief throughout the storm and inform the Operations Chief when the pump station pumps have either been throttled back or shut down. The procedure will reverse when canal levels lower and the Canal Captains will inform the pump station operators when canal levels are below Safe Canal Levels so they can adjust pumping accordingly. For steady flow conditions, the targeted level for the canals will be in the range from 3 to 4 ft. (NAVD88 Datum) this will provide desired suction head to maximize flow for Interim Control Structure Pumps while not approaching the Safe Canal Levels. Flow meters will provide discharge flow rate data for each Interim Canal Structure Pump. Canal Captains will coordinate with Local Partner Pump Stations to achieve steady flow under these conditions."

A few notes are necessary. First, let me explain the numbers on the "targeted level" in the above paragraph.

The only copy of the manual that has been released is from July of last year. At that time, the Safe Water Level was 5 feet at both 17th Street and London Avenue. Since then, the SWL has changed to 6 feet at 17th Street and 4 feet at London Avenue. Elsewhere in the manual, levels one foot below the safe water level are noted as the activation depths for the floodgate pumps. Thus for London Avenue, the steady flow operating range is probably 2 to 3 feet now, not 3 to 4 feet. And for 17th Street, the range might be 4 to 5 feet, or it might have stayed the same.

Also, the Canal Captains are the Corps Operations representatives in the S&WB pump stations. They'll be huddled over the readouts from the SCADA system, checking the canal depths from the level gauges and the flowmeter readings from the floodgate pumps. As one can see, the flowmeter readings will be part of the data considered by Corps personnel in their decisions to advise S&WB personnel to turn off the city's pumps.

For storms below category 3, Corps personnel are supposed to remain at the floodgates. Supposedly, they'll button themselves up in the little bunkers on the drive platforms, but considering the courage it takes to volunteer for such duty, I seriously doubt they will stay inside if there's a problem or if a request comes over the radio to check a flowmeter out on the pump platforms. Thus, there will likely be a backup for the flow signal received in the S&WB pump stations in the form of Corps Operations person who can look at the local display at the gates, as well as look at the pumps themselves.

Problems arise, however, when the storm is a category 3 or above. In that case, the operators are supposed to evacuate the floodgates. Now, I think it's questionable that they will do so. After all, these guys are probably pretty dedicated, and would most likely have to be escorted off the drive platforms at the point of a gun. I would compare that to the situation exemplified by the S&WB drainage pump operators who stayed during Katrina.

However, sometimes orders are orders, and you must leave. That situation is demonstrated by the evacuation of Jefferson Parish's pump operators during Katrina. With the operators gone, and with no automated system to control or adjust the pumps, the lake backflowed through the JP pump stations into most of the east bank of Jefferson Parish, a totally manmade disaster.

So for truly nasty storms (cat 3 and above), there might be no crews out at the floodgates. In addition (and this is according to the manual) for category 4 or 5 the Corps also pulls the Canal Captains out of the S&WB stations and relocates them to the Corps headquarters on Leake Ave at the river. The S&WB determines whether to leave the S&WB operators in the S&WB stations.

What this means is that all the data to determine whether to turn off more city pumps than will already be turned off (and they will be turned off after the gates drop; otherwise, they'd overwhelm the Corps pumps) will be received via the SCADA system. This includes the data from the flowmeters, which is most likely inaccurate, possibly greatly so.

So two questions need to be asked:
1) Assuming that the use of ultrasonic flowmeters was the correct choice in this application, does the Corps know how accurate the data from those flowmeters actually is? That is, has anyone done detailed baseline tests to determine whether there really is an error due to the current installation position of the pump flowmeters?
2) Does the Corps plan on running drills in which there are no operators interacting with the machinery at the floodgates, in order to determine whether there are kinks which need to be worked out? After all, their plans call for the floodgates to be evacuated in heavy storms.

Monday, May 21, 2007

Troubles with the trench again

The following is unconfirmed, but it's from a pretty damn good source.

During the Corps' gate closure drill on May 15th, 2007 which took place at all three floodgate sites, apparently they had some trouble closing the gates at 17th Street. Why? There was construction debris in the trenches into which the gates seat. Apparently, at least one and possibly more gates stopped about a foot short of actual closure.

There's lot's more information about the trenches at this earlier post.


With hurricane season approaching, I thought I would summarize all the problems that I've discovered with the floodgates pumps over the past few months. This list is somewhat scary.

1) Only one type of backflow prevention, instead of two as called for in Corps engineering manual for this type of installation. [blogged September 29, 2006]

2) Durst gearboxes undersized for anticipated thermal load, increasing potentiality of overheating. [blogged March 8, 2007]

3) Hydraulic reservoirs undersized by factor of three, increasing potentiality of overheating. [blogged May 11, 2007]

4) No electrical backup for diesel engines, despite availability from chosen vendor, decreasing reliability. [blogged May 11, 2007]

5) Undersized hydraulic fluid transmission lines, increasing potentiality for catastrophic failure. [blogged April 17, 2007]

6) Unpainted, rusting hydraulic fluid transmission lines, increasing potentiality for catastrophic failure. [blogged April 5, 2007]

7) Rusty hose fittings on pump units. Note that this one has supposedly been addressed by the replacement of most of the hoses and fittings at the pump units with hard piping. [blogged April 5, 2007]

8) Flowmeter sensors in locations which will amplify measurement errors [blogged May 22, 2007]

There are others that I just haven't gotten to as yet.

Friday, May 18, 2007

GAO, testing, and real problems

Yes, the GAO has presented their findings to Senator Landrieu's office. I'll have more to say about this later. However, I urgently need to zero in on the high-profile testing the Corps has been doing after refitting of the pump unit hydraulic motors. They are also replacing much (not all) of the hosing at the pump units with hard pipe. They are continuing to act in the exact bad-faith manner called out in the GAO report.

This is from the press release about the GAO report found on Senator Landrieu's website:
"By June 1, the GAO told Sen. Landrieu, the Corps plans to have completed reinstallation of 40 pump systems that have each been tested for 45 minutes to two hours."

Two hours?!? 45 minutes?!? Are they serious?

The most recent test, at Orleans Avenue on Tuesday, May 15, 2007, only went for less than an hour. Why is the media, our government officials - hell, anyone else but me - not noticing the tremendous disconnect between these dog-and-pony shows the Corps is putting on and what is really needed to determine if the pumps work?

These pumps and drive units may need to run for somewhere between 12 and 24 hours during a storm. Remember that the original factory testing called for full performance testing. Then that was downgraded to a five hour endurance-and-reliability (E/R) test for the drive units only, with no performance testing for the pumps themselves. Then that was downgraded to a three hour E/R test. And by the end of the testing (according to the complete batch of inspection reports I got via FOIA recently) , some drive units were being sent to the field after just 45 minutes of testing. Only a tiny handful of the pump units were verified as meeting the specifications for flow and head.

Shockingly, one of the findings in the GAO investigation is that the Corps has been making up testing requirements as they go along. It's shocking because it seems they continue to do so to this very day, and are actually using the latest batch of tests as a defense that they're now doing the right thing. It's up-is-down thinking, and so brazen in its attempt to flim-flam the public and policymakers that I can't believe it. Why are these systems not being run for legitimate amounts of time? How can we be sure that they will work as designed if they haven't been tested under the conditions they are likely to see?

I understand that due to canal levels, it's impossible to run the entire systems (i.e. five or six pumps at a time) simultaneously for an extended period. However, there is no excuse - NO EXCUSE - for not running each individual drive-and-pump combination for a real world length of time. The Corps is continuing - to this day - to leave us in jeopardy of pump failure.

Wednesday, May 16, 2007

The walls

I've been thinking a lot about the big picture, and how we've gotten to the atrocious situation we're now in for flood protection. We now have a set of conditions which are more likely to cause flooding in New Orleans than before Katrina. How did this happen?

One part is the Corps' New Orleans District's moronic decision - yes, decision - to not fix the outfall canal walls, instead putting up the floodgates. The walls were supposed to protect against storm surge, and they were only put up within the last fifteen years. Why could they not be made to work now? It's hardly rocket science. After all, since Katrina the Corps has reinforced or improved walls and levees along the Duncan Canal in Kenner, along the Gulf Intracoastal Waterway by the NASA-Michoud site, and at plenty of other places around the region. Why were the walls along the outfall canals any different?

I'm really not sure of the "why." But I am sure of where we are: no wall improvements in the near future, and possibly ever. This has deep consequences, some of which we've already seen (pumps at PS#3 getting turned off twice in the last five months for one). Nearly everything that is a problem with post-Katrina drainage and flood protection in New Orleans (on the east bank, west of the industrial canal) can be traced back to the Corps' foolish, foolish pair of decisions to go with the floodgates, and to not repair the walls. This post talks about the consequences of, and possible reasoning behind, those decisions, and where we are as a result.

Let's start with evidence backing up my premise that the walls won't be repaired...

Evidence the walls will not be touched
Of course, the plain evidence (which the Corps always tells us to ignore, but which my gut tells me is right) is right there. There has been no work on the outfall canal walls outside of the breached and heavily damaged areas since the storm. Whenever I tell this to someone from outside the region, they seem amazed.

There was also this April 25, 2007 story from WWL-TV, in which Corps Protection and Restoration Office Chief Tom Podany is quoted as telling the Jefferson Parish council that the west walls of the 17th Street canal "meet all the requirements we have." Remarkably (or perhaps not, when you consider the sub-par caliber of their recent Corps coverage), the Times-Picayune devoted nothing to this story until Senator Vitter let the Corps know such statements are way out of bounds, since the canal walls are part of the Lake Pontchartrain and Vicinity project, and are thus required to work as designed, no matter what malarkey the Corps builds at the lakefront. His letter can be found here, and the Times-Picayune's story is here.

Then there's the other story from the April 13, 2007 Times-Picayune about the west walls on the 17th Street canal, in which the Corps "revealed" that the sheet piles on the west side are very short, but still expressed strong confidence in their floodworthiness. I put revealed in quotes, because that particular article is so poorly composed, and full of such inconsistencies, that I am convinced it should have never been written in the first place. It will be subject of a future post about what happens when bad stories happen to worse writers. But it's still helpful in revealing the Corps' feelings about the walls, which are that the walls are just fine. The April 13th article also mentioned the delivery of the Corps' study of the walls' stability by the end of April. We're still waiting. Finally that article mentioned finishing emergency repairs on the east side of the 17th Street canal by June 1. Those repairs were first announced April 3, 2007, in another Times-Picayune article which included the usual head-patting from the Corps.

That work just started in the last few days, but my reaction is: Huh? I thought the walls were fine. That's what we've been told for over a year. Another reaction is: why did it take over 21 months, if the job can apparently be accomplished in two weeks (I sincerely doubt this is a comprehensive fix, especially since the area covered is so small), and what does that say about the condition of the walls in those spots up until now? It seems the Corps is talking out both sides of its mouth, which certainly doesn't breed trust. At the very least, there's been a serious lack of urgency. Why wasn't this work done before last hurricane season?

[Update, 7/25/07]
As far as the present condition of the walls along the 17th Street canal, we can now know a bit more about that.

On May 24, 2007, the Times-Picayune reported on the release of a Corps study of the 17th Street canal. The study looked at how stable the walls are under various pumping and lake stage scenarios. They looked at the envelope between practically no flow to maximum output from all three pump stations (PS#6, I-10, and Jefferson Parish's tiny Canal St. station). They also looked at the envelope of the lake level from 0 feet to 5 feet, in one foot increments.

They found that the six foot safe water level can be justified, though the water rises dangerously close to it in certain regions of the canal under many circumstances. Those regions (near I-10 and Veterans) are the ones that got the emergency repairs mentioned in the April 3, 2007 article.

Well, I have received a large chunk of that report, and now everyone can read it for themselves. You'll find the first 58 pages of the report here. The Appendix, which contains the raw data, a May 2006 inspection report, a fascinating November 2005 Linfield Hunter & Junius internal memo, and more good stuff is here (warning: over 80 megs). The rest, which includes the actual analysis, should be forthcoming soon.

[Update within update, 6/1/2011
The complete report, including the back matter like the graphical output of the analyses, was placed on the Corps' website as part of the IER#5 public comment period in June, 2009. It is still there. But just in case it disappears, here it is for posterity.]

There's lots to take away from the report. Here's a few of my impressions.

1) The sections of the east wall immediately adjacent to the breach are among the weakest in the whole canal.

2) The Corps' method of calculation for all the governing Safe Water Levels (the Method of Planes) was superceded by Corps HQ shortly before the study was issued by something called the Spencer Method. They say they will go back and redo the calc's using the newer authorized method, but there's no indication whether it would give different results. I'm no geotechnical engineer, so I have no idea whether this is significant or not.

3) There's no correlation between the various "reaches" into which the canal walls were broken for calculation purposes (there were about 16 on each side) and the placement of the depth sensors which are tied into the SCADA system (I think there's only five or six of them). I guess the Corps operations folks just watch the particular sensor nearest the weakest spots.

4) With the lake at five feet, the water gets above or gets within six inches of the governing safe water level near I-10 for total flows exceeding 7500 cfs. This means if the lake level gets close to five feet and the gates haven't been dropped, PS#6 will have to cut back on three of their big pumps. Of course, that's kind of moot, since there's not going to be more than 7500 cfs available with the gates down anyway (and most likely a lot less, due to failures of the MWI pumps).

With lake stages lower than five feet, it appears there is no significant need for pumping cutbacks at PS#6. However, the trigger for gate closure is six feet in the lake. And also, as noted above, these calculations are based on a now-outdated method. In addition, they rely on the Corps' ridiculously low safety factors. Combine all that, and there's more than enough uncertainty to make one believe that pumps will be cut back in extreme cases, whether the gates are up or down.

Overall, however, the report appears to present mostly good news. However, it's important to note that there was no outside review of this report before its compilation, and there were no outside resources utilized by the Corps. This is unfortunately still the norm.

However, now that the report is public, researchers and scientists outside the Corps can now critique it.

[end update]

So it's pretty clear that the Corps New Orleans District, which has had 21 months to do something with the walls, has chosen to do nothing (excepting some minor patching). Why? According to them, it's because the floodgates are there, and they "take the walls out of the equation."

Out of the equation?
When asked about the walls, the Corps says, "Don't worry, the floodgates take the walls out of the equation." Podany said it in that Jefferson Parish Council meeting:
"'Because we have that temporary gate at the lake, those walls don’t need to function during a storm,' said Podany."

They said it in the April 13th T-P article:
"'We've done the analyses, and the results convince us that underseepage wouldn't cause the walls to fail anywhere in the 17th Street Canal because of the steps we're taking, which includes limiting the amount of water in the canal,' said Walter Baumy, chief engineer of the corps' New Orleans district."

They said the same thing as recently as May 9 in the Times-Picayune, at the end of the long overdue article about everything the Corps has done wrong since Katrina:
"Despite the weak spots, Bedey said, the corps has been successful in dramatically reducing the risk from storm surge in a number of locations, compared with before Katrina. That includes the installation of gates on the 17th Street and London and Orleans avenue canals, which will eliminate the threat to flood walls along the canals. Problem pumps designed to move rainwater in the canals over the gates into Lake Pontchartrain have been repaired and successfully tested at the 17th Street and London Avenue canals, Bedey said, and are expected to be repaired and tested at the Orleans Avenue canal by June 1."

That's nuts. All these statements rest on the giant, unstated assumption that the gates and floodgate pumps will work as designed. To no one's surprise, I have serious doubts about that assumption, and so do many within the Corps. As I pointed out in my last post, if the gates don't seat properly in the trenches, they'll probably fail. If the gates' hydraulic winching mechanism doesn't work, and the winds are too high for a crane (i.e. over 30 mph), they won't close. There are plenty of ways for this system to fail. We also know that the MWI pumps and drive units have deep problems.

In addition, the walls see water all the time from normal rainstorms when the gates aren't down. The gates most definitely do not take the walls out of the equation, unless the equation only counts tropical storms in which the gates don't fail (more about gate failure below). Otherwise, the walls are very much in the equation. This knocks out a central premise in the construction of the floodgates, forcing the Corps to retreat...

The Corps has a "system" to save the walls, and drown us
When the fallacy of the "walls are out of the equation" is pointed out, then the Corps says, "Don't worry, we have a system to keep the walls from seeing too much water during a rainstorm." In point of fact, they have no such system. They have a system for monitoring the water level in the outfall canals during a rainstorm, which is an entirely different kettle of fish. That system is a set of electronic level gauges installed along each canal which is connected to a few computer terminals scattered around the city, including within the three pump stations at the southern end of each outfall canal. Surprisingly, the T-P actually described the system correctly in their sloppy wet kiss to the Corps on March 25, 2007 (keep in mind, the following describes what happens with the gates closed, not during a non-tropical system):
"...The canals themselves have been lined with a Supervisory Control and Data Acquisition, or SCADA, system that uses sensors to transmit a wealth of information via computer screens, including changes in water levels along the length of all canals.

It is the SCADA [system screens] inside the pump stations that canal captains will monitor as they synchronize the pumps near the gates and the bigger city pumps situated hundreds of yards -- sometimes a mile or more -- closer to the heart of the city.

With the gates closed against a surge, city operators must power down their stronger pumps to keep from overwhelming the lower capacity of the dozens of pumps set up at the mouths of the outfall canals. It is the pump operators who will make adjustments in response to information the canal captains collect from SCADA readings, Wagenaar said.

So critical is SCADA to protection of the canals that there are special backup generators in the pump houses just to keep the data system running. In addition, fiber optic cables backed up by a microwave system have been installed to keep SCADA data flowing as long as possible to pump stations and to the corps' riverfront headquarters and emergency bunker."

What's notable is that - other than this article - it is rare that the distinction between monitoring and control is actually made, especially when the Corps' New Orleans District is quoted directly. It is an important distinction, and its omission is typical of the self-aggrandizing puffery the N.O. District has adopted as an unbreakable habit. The Corps doesn't control anything outside the gates, but they act like they do.

During any rainstorm, including non-tropical ones, if the water level gets too high in any of the outfall canals as it is being pumped out of the city (that is, above the Safe Water Level - or SWL - for each canal: eight feet at Orleans Avenue, six feet at 17th Street, or a measly four feet at London Avenue), then there are only two solutions to avoid another breach:

1) Lower the level of water the city's pumps must push against from the lake. That is, the Corps shuts the floodgates and turns on the floodgate pumps. God help us.
2) Limit the amount of water going in the canals. That is, the Corps tells the Sewerage & Water Board, a local agency over which they have no official jurisdiction, to turn off the city's drainage pumps for the specific canal for which the SWL has been exceeded. This has already happened twice, on December 30, 2006 and May 4, 2007.

That's the "system." Force the city to drown itself to save the walls. Talk about bass-ackwards. So the unfixed walls are not only a problem not remedied by the floodgates, their unrepaired state is actually driving decisions on flood protection. The walls still loom very large.

What if the gates fail?
In choosing to spend over $200 million in construction, repairing, changing, and expansion of the floodgates and pumps, the Corps deliberately left fixing the walls in some fuzzy future. They cast their entire lot with the floodgates, as the quotes above show. But what if the gates don't work?

Such an idea - despite all the evidence that the New Orleans District and its mostly locally-based consultants can't be trusted to design a block of wood - still seems anathema to almost everyone. But think about it. This same crew was trusted for 40 years to do the right thing, to follow their own rules, and they failed to do so. They even decided - decided! - to construct the levees and floodwalls to improper benchmarks, resulting in shortfalls of two feet when one compares actual heights to design heights! Now, we are supposed to believe that structures conceived and designed during the same time the Corps was denying responsibility for Katrina's devastation (from September to December of 2005) are going to be stronger than anything else they've ever put out there? We're supposed to believe that - despite a documented history of non-functionality for over a year - the MWI pumps are all of sudden going to be better than sliced bread? We're supposed to believe that the same bunch that allowed the trenches at the gates to be filled and sealed with two feet of grout during the height of the 2006 hurricane season somehow now knows their ass from their elbow? No, there's been too much wrong with all of this to believe it can work. And if the gates don't work, then we're back to relying on the walls - the same walls that are known to be underdesigned and poorly constructed.

So what if all that money had instead been applied to fixing the walls? Could it have been possible? Would New Orleans have been safe during 2006? Maybe, maybe not. But to not even consider fixing errors which have been reported very widely and have been analyzed by at least three teams (IPET, Cal-Berkeley, Team Louisiana), is absolutely nuts. Heck, Team Louisiana's report even tells the Corps exactly how deep the sheet piling needs to be along London Avenue. The design work is already done.

How the floodgates - combined with New Orleans District incompetence - have made things worse in new and innovative ways
There are other real consequences to these really bad decisions. Let's look at one in particular.

In the floodgate scheme, there is not enough pumping capacity at London Avenue and 17th Street to match the pre-Katrina drainage capacity (approximately 9300 cfs on 17th St and between 7000 and 8000 cfs on London Ave.). There never has been, and I don't think there ever will be. Let's look at that situation in a little more detail.

My assertion that there probably never will be enough pumping capacity doesn't come from thin air. The Corps says as much in Rev. 4 of the PIR, issued in October of 2006:
"Based on a cursory review by ERDC [the Corps' drainage gurus in Vicksburg], their belief is that large discharge capacities required at 17th Street and London Avenue may be difficult to achieve. This concern is based on a large number of pumps installed in a very small canal area with an erratic distribution of flow. A definite recommendation cannot be made at this time by ERDC."

I'd say we know the recommendation now, seven months after issuance of that report. The plain evidence is this: the Corps is spending $135 million to get pumping capacities up to levels that are still less than pre-storm values (about 7500 cfs at 17th St, and about 5000 cfs at London Ave). A lot of that money had to be routed through an extraordinary Air Force contracting instrument outside of normal Corps contracting procedures. Does anyone really think a similar, uniquely funded sum will be spent any time soon, considering how much has already been thrown down this hole? I'd say "that's all folks" on extra pumps. We're stuck with what we've got coming this summer, assuming they all work.

In addition, we've also got evidence that they've also unofficially put the kibosh on more pumps at 17th Street, as mentioned in a little noticed article in the Times-Picayune on May 6, 2007.

The article was about the townhouses which line the west side of the 17th Street canal floodgate site. They are collectively known as Mariner's Cove West, and are part of the Mariner's Cove development, which went up in the 1970's. Before Katrina, they were very valuable properties.

The Times-Picayune missed the entire point of the article, focusing on the homeowners and their high incomes. But Mariner's Cove West has been key to this entire pumps-and-gates story, without anyone noticing since the beginning.

Mariner's Cove West has actually been used by the Corps for over a year as an excuse NOT to expand pumping capacity at 17th Street. In Rev. 2 of the PIR, issued in May of 2006, the Corps laid out their case for adding about 4000 cfs of additional pumping flow at the 17th Street floodgates. Most of that flow in the Rev. 2 scheme (which was eventually mostly scrapped) would come from hydraulically-powered pumps along the west side of the canal. Why not put more pumps in along the east side?
"Adding capacity beyond the additional 4,000 cfs would require considerable additional rights-of-way. On the east side, sub-alternatives are limited by townhouses."

Yeah, they were afraid to tear down Mariner's Cove West. This despite every owner of every unit wanting a buyout since right after the storm. But very early on, the Corps New Orleans District (specifically Fred Young) told Mariner's Cove West townhome owners that the Corps had no intention buying them out. They were told this as early as February, 2006 by Mr. Young, and told the same thing many times after that. The Times-Picayune did a big story about Mariner's Cove West on April 19, 2007 (about a year late, in my estimation), mostly focusing on the struggles the townhouse owners have faced with the Corps' N.O. District Real Estate Division.

Let me make clear exactly what I'm saying here, because this is new. The Corps, knowing they needed more pumping capacity at 17th Street, refused to make way for that capacity - even though every person who would lose their home was willing to be bought out. Instead, the Corps dithered for over a year, leaving the Mariner's Cove West folks in limbo, and - more importantly - leaving everyone who drains into 17th Street without their prestorm pumping capacity.

Now, back to that May 6th Times-Picayune article, because it gets worse. It's about the Corps finally buying out Mariner's Cove West. Except - and you're not going to believe this - they're not going to use the land for anything! Here's the key quote from the article:
"'I just wanted out, but this helps others, who were unsure, to move on,' said [Mariner's Cove West resident Mike] Frank, who has left his native New Orleans for northern Florida. 'They're telling us none of the property will be a project; they will make it green space, so that's good news for the rest of Mariners Cove.'

I nearly spit up my drink when I read this in the paper the day it was printed. I also was amazed that the T-P didn't make the connection between pumping capacity and Mariner's Cove. The folks at New Orleans CityBusiness missed it too when they reported the same basic facts, but with a little more depth in a May 14, 2007 article. They wrote one sentence about the use of the property and also got a vague quote from Col. Wagenaar:
"The Corps said it was necessary to acquire and demolish the buildings to continue current work and provide maintenance space for future projects. 'Because of the ongoing and future work at this location we made the decision to acquire this property,' said Col. Richard Wagenaar, commander of the Corps' New Orleans District."

You have to read the tea leaves and between the lines with these folks. Sometimes, what isn't said is way more important than what is said. And what isn't said - between the two articles - is that the Corps is buying the property in order to fit more pumps into the 17th Street site. That's significant. Do we really think they would hold that information back? Combine it with a) the already stated objections by the local Corps folks that the site is too small; b) what I know the Corps has been telling homeowners for the last year (yes, I've talked to many of those homeowners myself); and c) ERDC's worries, and I only see one conclusion.

So here's the connection between the Mariner's Cove West buyout (specifically its timing) and 17th Street pumps, for those too dim at the paper to make it themselves: there will be no more pumps at 17th Street, despite a bunch of real estate opening up at the site. We're not getting back to pre-Katrina pumping levels for a very long time.

So that's another consequence of placement of the floodgates in the first place: they couldn't actually provide a system that has the same level of protection as we had before Katrina. And in not repairing the walls and enstating the Safe Water Levels, they've constrained that level of protection even further, to the point where we're affected even when there is no hurricane. The two decisions - building the floodgates and not fixing the walls - have had deeply troubling consequences for our city.

The really big consequence - Flood Insurance
Last April, everyone in town was in a tizzy because FEMA had yet to release its post-storm guidance on flood insurance and BFE's (base flood elevations). The ABFE (Advisory Base Flood Elevations) for Orleans Parish was finally released on April 12, 2006. These have allowed the rebuilding to proceed by allowing residents and businesses to purchase flood insurance.

The ABFE's (they were issued for all Katrina- and Rita-affected parishes and counties), which were legally adopted by Orleans Parish on August 3, 2006, are conditional documents. That is, they are based on the Corps doing certain things by certain times. Specifically...
"In addition to the recent USACE storm surge modeling, FEMA has also developed these recommendations based on the height and integrity of the levee system expected to be in place by September 2007."

So what if the levee system - which includes the floodgates and pumps - is not what FEMA expected to be in place by September, 2007?
"Although FEMA is confident in the results from this current assessment, the agency will continue to monitor progress made with regard to levee improvements, findings from other ongoing studies, and enhancements to the agency’s understanding of the probability of flooding in this area. FEMA will adjust the recommended flood elevations as necessary as the agency prepares updated FIRMs [Flood Insurance Rating Maps, popularly known as "flood maps"] for Orleans Parish and its incorporated areas."

And what does that mean? You only have to go one paragraph backwards to find out...
"For the Parish Advisory BFE (ABFE) inside levees, this Guidance is similar to NFIP rules for areas protected by levees being restored to provide 1-percent-annual-chance base flood protection. Should the requirements needed for application of these rules fail to materialize, flood elevations in this area would be based on a "without levee" scenario and could exceed elevations of 8 feet (west and south of Mississippi River) or 13 to 14 feet (east and north of Mississippi River) referenced to the National Geodetic Vertical Datum of 1929 (NGVD29)."

Now a little explanation is in order. Last year, the Corps was saying they would have a certain level of work done by September, 2007. According this Corps press release, issued the same day as the Orleans ABFE, that level was,
"September 2007 - Completion of restoration of undamaged and subsided areas; completion of previously unconstructed portions of authorized projects"

It's unclear if that's going to happen by this September. There are so many projects included, and I don't know the status of every one. It's also unclear if FEMA is accounting for the floodgates and unrepaired walls in their calculations.

Restoration of the system to the 100-year level authorized by Congress was supposed to be done by 2010. However, the Corps has pushed some of that work out to at least 2012. It is that work which is supposed to allow for true FEMA FIRM's to be issued. But since the region can't wait until 2010 or 2012 or whatever for final maps, FEMA will work off of the Corps long-delayed risk and reliability study, which will supposedly include all the 2010 work.

My question is this: if the current gates and pumps and the lack of repair to the walls are shown to be failures, what happens to flood insurance in the city the day after that failure is confirmed? That is, will FEMA follow up on their statement in the ABFE to "adjust" the flood maps, possibly adjusting them to a "without levee" condition?

I'm certainly no expert on all this flood map stuff, but I can read. And I think the (admittedly very hard) question needs to be asked of FEMA's mitigation folks and the Corps, although we already know the Corps' answer - "everything's fine!"

It needs to be asked because everything in the recovery is based on the availability of flood insurance, and I mean everything. We're talking about billions of dollars in Road Home funds, as well as everyone's mortgages, which require flood insurance, all of which could be placed in jeopardy if the gates don't work, simply because the walls haven't been touched.

Conclusion - the Corps has royally screwed up
None of this would even be happening if they hadn't put in those stupid gates in the first place cutting off New Orleans' drainage capacity. But then they decided to not do anything to the walls, doubling down on their earlier stupidity. So they've made things worse for the past hurricane season, for the coming hurricane season, and for every season until 2013. That's a total of seven hurricane seasons, never mind the problems with normal rainstorms and the safe water level limitations. Like their actions with the datums and the sheet pile lengths (well chronicled in Team Louisiana's report, specifically chapter 5 and chapter 6), the Corps has deliberately bungled their mission of protecting the citizens of New Orleans.

The Corps will not fix the walls. They will rely on their beloved yet untested, unproven, and defective floodgates (we won't really know they work until a storm hits), leaving New Orleans in a more perilous condition than before Katrina. Residents should think long and hard about relying on the gates, the MWI pumps, and the unrepaired walls to save them. The situation has become unbearable, and I hope people realize it before the worst happens, rather than when it's too late.

Saturday, May 12, 2007


Updated quite often. See below at various spots.

In light of some very cloying, cowardly coverage by the local media...

a) WWL-TV does a report on London Avenue canal testing on May 10, 2006, and only interviews Corps personnel, and then shows the wrong spot on the canal where the test will take place [they showed the repaired breaches at Robert E. Lee Ave, which is over a mile north of the area where the tests will really take place - just south of Mirabeau Ave.]!

[Update, 6/1/07: I have issued a correction for the above paragraph, which is erroneous. The correction can be found here.]

b) WWL-TV does a special series of reports titled "Closing the Gaps," from April 30th to May 3, 2007, supposedly about all the problems we still face with levees and pumping, problems brought about by the Corps. So who do they interview as the main source of information... the Corps! Part 1 here (video here), part 2 here, part 3 here, I couldn't find the print version of part 4, but the video is here.

c) On May 9, 2007, the Times Picayune republishes a Corps press release about the new head of the New Orleans District as an "article" without mentioning that the current commander - Colonel Wagenaar - is retiring a year before his usual 3 year tour is up.

d) This one reaches back a bit, but on Sunday March 25, 2007 the Times-Picayune published what was portrayed as a "big" front page article, a real love letter to the Corps, detailing how ready the Corps was to save us all. In reality, it was just a huge rewrite of the Floodgates Operating Manual, a document which has been publicly available here - due to my requests - since last August. Also the article included only quotes from Corps of Engineers personnel (a hallmark of a lot of Times-Picayune coverage of the Corps these days), including a tremendously cheap shot by Wagenaar:

"...And although Wagenaar applauds the 'heroic' work of local emergency responders to the 17th Street Canal breach -- especially West Jefferson Levee District employees who engaged in road building -- he said he welcomes the streamlined chain of command achieved through levee district consolidation. 'What is not streamlined is dealing with the Sewerage & Water Board. They don't seem to work for anybody.' Wagenaar said."

Of course, the article's author leaves that comment - which is the height of hubris from an agency that seems to answer to no one, including Congress - unrebutted. She apparently didn't even bother to call the S&WB for comment. Also missed in all the glorification of the Corps' communications plan is this actual quote from the Manual:

"One Corps of Engineers radio will be furnished to the S&WB for communications in addition to having the Canal Captains stationed at their respective S&WB pump stations."

Yep, that's right: the agency in charge of pumping gets all the benefits of a single radio to communicate with the Corps.

[Update, 6/7/07: The radio has been addressed as part of the kerfluffle over control of Drainage Pumping Station #6 on the 17th Street Canal. See my June 7th post here.]

... So in light of all that (and much, much more) crappy coverage of what the Corps has done and isn't doing over the last two weeks, allow me to let you in on some real stories that anyone with a modicum of investigative sense would have ferreted out by now:

1) The hydraulic fluid reservoirs on MWI's engine skids are undersized by a factor of 3. MWI's own catalog calls for a 900 gallon reservoir. FPI's 60 inch pumps use a 900 gallon reservoir. The MWI models at the floodgates have 300 gallon reservoir. This decreases the capacity to get rid of heat generated when the pressure of the hydraulic oil is raised up to 3000 psi (theoretically - though it's probably closer to 2600 psi), and then shares some of that energy with the Rineer hydraulic motor in the pump unit. If there's too little room in the reservoir, the fluid doesn't have enough time to cool, meaning the oil stays thinner than it needs to be. Less viscosity means less ability to transfer energy means less ability to pump water.

Why did MWI do this? Maybe to save money on hydraulic fluid, which they were responsible for supplying for startup and commissioning. Such a savings could have been in the range of $50,000 to $150,000, money that goes straight to MWI instead of some supplier.

According to my latest batch of testing reports, received via FOIA request (and held back by New Orleans since January), during testing MWI actually was pointing fans at the drive units to keep them cool until the Corps QA's told them to stop it. Overheating of many components - not just the hydraulic system, but also the Durst gearboxes - on the drive units is still a big, unaddressed concern.

2) The Corps has yet to issue three Orleans Parish pump station repair contracts, over 20 months since Katrina. These contracts are part of the original $40 million set aside for pump station repairs. The three contracts - Electrical-Mechanical Repairs, Building/Structural Repairs, and cleanout of the PS#17 Discharge line and the inlet basins of three or four (it's unclear if Elaine Street's basin cleanout is included in this contract) other stations - appear to be on the very slowest track possible, since the Corps has had the money since January of 2006. The following is a quote from a February email sent to me by one of the people working on the Discharge Line solicitation (W912P8-07-R-0027), which has been indefinitely postponed since February 12, 2007:

"The reason we suspended the solicitation was due to the Right-of-Entry agreements we are working on getting from the Port of New Orleans and Project Permits from the Orleans Levee District and DOTD. We were under the assumption that we could do the work under the ROE and CA with the New Orleans S&WB. We are doing this so that everyone knows what is going on and we don't have any delays once the project is awarded. The ROE from the Port of New Orleans has been obtained and we are working on getting the two project permits."

Sure sounds like urgency to me!

[Update, 7/9/07] And the ugency continues on this one. The PS-17 Discharge Line/PS-6 & PS-7 Basin Cleanout solicitation was revived on June 5th, and has since gone through numerous changes and delays, still without award of a contract. The latest modification (number Seven), issued July 9th, changes the response date to July 11.

Also, one of the other remaining Orleans Parish repair contracts was finally issued a couple of weeks ago. According to this Corps contract award announcement, a $3.25 million contract for Electrical/Mechanical Repairs at the S&WB drainage pump stations has been awarded to Healtheon, Inc on June 19, 2007. These repairs will happen at nearly every one of the 23 stations.

While the award announcement gives a local addess (Suite 2500 in the downtown skyscraper at 201 St. Charles Ave.), the Louisiana Corporations database shows that Healtheon's principal office is in Birmingham, Alabama. A Google Maps search of that address shows it to be a private house. Its webpage shows its focus to be in "healthcare construction," but it appears to have a division which just happens to deal in "disaster recovery." Healtheon does not appear to be registered as a contractor in Louisiana, according to the Louisiana Licensing Board for Contractors website.

Healtheon also received the contract for Electrical-Mechanical repairs to pump stations in St. Bernard Parish, as noted in this January 28, 2007 award announcement. The value of that contract was about $786,000. How is it there was only one company anywhere that could handle these two contracts? Why wasn't this open to competitive bidding?

The Orleans Parish E-M repair contract, like the St. Bernard Parish E-M repair contract, was yet another sole source 8(a) set-aside. Sole source 8(a) awards have become the pattern for the Orleans Parish pump station repair contracts. Almost all of them, except the motor rewinding contracts, have been sole source set-asides. There's no competition required, and I've yet to find where one can even find solicitations for them. They stink to high heaven. And what the heck took so long to issue this latest contract? I'll tell you - no urgency.

Speaking of which, two other Corps pump station repair contracts still remain unissued. They are the Building & Structural Repairs contract, and the Elaine Street Station Reconstruction contract. The Elaine Street Station reconstruction contract is somewhat new, having appeared on the docket within the first few months of 2007.

[Update, 7/24/07]
As I surmised above, the PS-17 Discharge Line/PS-6 & PS-7 Basin Cleanout, the Building & Structural Repairs, and Elaine Street Reconstruction contracts are still unawarded, according to the July 19, 2007 schedule of Orleans Parish Pump Station Repairs from the Corps.

Also extremely disappointing embarrassing is the fact that the repairs are moving at a snail's pace. As we approach the two year anniversary of the storm - and as we just passed the 18-month anniversary of the signing of the bill which appropriated the money for these contracts (the 3rd Katrina supplemental was signed December 30, 2005) - three contracts remain unawarded and fewer than half of the 14 contracts have been completed. In addition, those that have been awarded are now taking even longer than previously scheduled.

For example, contract OPS-6A, (awarded to IPS of Louisiana and then subcontracted to Conhagen; Corps award notice from 12/1/2006 here), intended to replace bearings in vertical pumps at five pump stations, was previously scheduled to take 243 days, according to the Corps' November 1, 2006 repair schedule. The latest schedule shows that work now taking 290 days and dragging into the end of September, 2007. And if one takes a close look at the percent completion (42%) versus the percent of time passed since the contract started (75% of the anticipated 290 scheduled days) it appears that the work on this contract is even further behind than indicated. I doubt the contractor will be able to finish 58% of the work in 25% of the time. The delays seem to be concentrated in the three New Orleans East stations (14 and 16 on the lakefront, and 15 way out east near the end of the world), where construction progress is respectively 19%, 12%, and 26%.

Looking at the big picture, the scheduled completion date for all of the repair work is February 9, 2008. That's 34 months after the storm hit!

They really don't care, do they?

3) [Update, 5/16/07] Has anyone asked about the Corps' plans to evacuate workers from the floodgates for anything over category 2 (see the Floodgates Operating Manual)?

Also see this 5/16/07 article from the Times-Picayune:

"The other three members of each corps gate team will stay at the gate for storms up to Category 2 hurricane strength. For Category 3 storms, they'll relocate to a safer structure nearby. For Category 4 and 5 storms, considered worst-case hurricanes, one three-man team will retreat to a bunker at the corps headquarters complex on Leake Avenue at Prytania Street, while the other two teams will evacuate the area until the storm passes, Wagenaar said."

According to the manual, that nearby safer structure is the Louisiana Dep't of Transportation & Development building at 7252 Lakeshore Drive, shown here. It's about a quarter mile from the 17th Street gates, which makes it quite far from the gates at Orleans and London.

There's no way to remotely control (other than on/off) or adjust the floodgate pumps; that has to be done at the drive skid. The pumps will very likely need manual adjustments - which will be impossible with no one there. This was buried at the bottom of the T-P March 25 article, written by a different, less rigorous author than the May 16th article, and made to sound not very important:

"If the area is threatened by a Category 4 or 5 storm, the men probably would be pulled off site once the gates were down and the pumps running. They would return as soon as possible after the storm passed, Accardo said. 'Why have them in harm's way? These aren't really safe houses in the sense that Jefferson Parish is building safe houses,' he said. 'These are equipment houses designed to handle a hurricane, but not a major storm surge.'"

Well, what if there's a problem with the fuel lines from the 20,000 gallon diesel fuel tanks? By the way, the Corps could have gotten dual drive units, with electrical motors as well as diesel ones - MWI offers them in their catalog, which is the basis for the Corps specs after all. Or they could have specified a drive system that could actually be controlled remotely, rather than one that can barely be turned on and off by remote. But why bother going for backup power or a truly robust control design? They were in a hurry!

In any case, the Corps knows leaving the pumps alone is a problem, but they just skirt over it in their own manual. From "Step 8 (T - 12 hours)" on page 10 (Adobe page 14):

"Due consideration must be given to what storm conditions are expected to be after the gates are installed and the pumps are initialized to permit the personnel that are part of the closure operation will be out of harm’s way as they travel to local safe locations or are being evacuated to remote safe locations."

That is, the Corps plans on "initializing" the pumps, i.e. turning them on and hopefully maximizing their flow, and then leaving them. What happens if they shut down and can't be restarted remotely (a very likely circumstance if one reads how to start them in the first place- which definitely requires someone being physically in front of the drive unit)? Well, I guess then we're out of luck. The Corps says, "c'est la vie."

I wonder if they plan on rehearsing leaving the pumps unattended the entire time they're running in their drills for this hurricane season? I sincerely doubt it.

4) Does the entire city know and understand that we're depending on a contract to a diving firm to guarantee the gates actually work? When the gates are lowered, they seat in a trench, which a firm called H.J. Merrihue is responsible for keeping clear during operations. If the gates don't seat in that trench, they'll fail. I'd like to know exactly how the diving firms do that clearing while a tropical storm is coming into the lake, and the city is pumping rainwater out along the canals. Plus, the Corps says the diving firm is responsible for evacuating their own personnel after the gates are lowered, despite the Corps getting the Corps folks out on helicopters under a separate contract. Again, see the Floodgates Operating Manual for the details on this one.

These are the same trenches that the Corps' contractors filled with grout during the height of hurricane season last summer without realizing it. They had to chip the stuff out mechanically at 17th Street and Orleans Avenue, meaning had they tried to lower the gates before the trenches were clear, they wouldn't have seated, and the city would have been defenseless against storm surge. I broke this story last August and posted about it to a number of local newsgroups. You can find my postings here, here, and here. I also posted some post-game analysis, including catching the Corps in some lies about the importance of the trenches, here. My concerns about the gates seating still exist, and have been echoed in National Geographic's latest feature, from which comes this nugget about the floodgates:

"A Dutch engineer recently visited some of the new floodgates and pumps installed at the mouths of the city's three main drainage canals. His verdict: They may be 'doomed to fail' in the next big storm. The engineer, who asked not to be named because he sometimes collaborates with the corps, notes that the gates have no mechanism to remove sediment and other debris that might keep them from closing as a storm approaches. Instead, the corps says it will rely on divers to check for obstructions and clear them away."

This is very troubling. Supposedly, the Corps was supposed to construct some metal pieces designed to fit in the trenches to keep them clear until they're needed, but of course we never heard whether that happened, or if it did, if that idea was successful. Plus, I'm pretty sure the trenches themselves are pretty beat up from all the jackhammering last summer, so I doubt they could build pieces that would completely seal against silt, rocks, and other debris building up. It's just another example of how what appears to be robust could in fact be very weak.

[Update, 7/24/07]
The grout in the trench problem was apparently very serious. It was noted in a PowerPoint slideshow given June 27, 2007 by the consulting engineers on the floodgates job at last months' SAME Infrastructure conference. The main conference webpage is here, and the presentation on the gates - given by guys from Linfield Hunter Junius and URS - is here.

Here's the exact text regarding the grout problem, from slide 43. Note the final bullet point.
"- Soil Mixing
• Soil Improved to an Average Shear Strength of 1,000 PSF minimum
• Allowed Fewer Piles
• Critical Item in Construction Schedule
• Soil Mix Residue Issues – Grout Silted in Gate Foundation and Pump Sumps – Major effort to remove soil mix residue with divers"

There are also pictures of the grout that flowed all over the site and is still there (slides 73 and 74). The grout went everywhere. It looks like someone stepped on a giant tube of toothpaste.

The presentation is fascinating, since it contains information never before revealed, like alternatives considered to the final needle gate design. One choice was to sink barges!

The presentation appears to have been thrown together fairly hastily; the pictures of the grout from 17th Street are actually dated two days before the slideshow.

And of course, they are still putting lies about the pumps out there. There is a mention of them on slide 47: "Hydraulic Pumps – Only Type of Pump Available in Allotted Time." This fails to mention that the pump specification was taken straight out of MWI's catalog. It also fails to mention all the problems before and since the installation of the pumps.

One other thing that isn't mentioned: the sill at London Avenue was completely screwed up and delayed startup of that floodgate for over a month. The sill problems were mentioned in Times-Picayune in a May 12, 2006 article (back when the T-P actually cared what was going on around them). There were more details in this July 3, 2006 article from something called "Construction News:"
"Complications with the sill prevented M.R. Pittman from completing the London Avenue Closure by the June 1 deadline. 'We expected to have the sill complete, but we ran into difficulties with the sheetpiling,' said Leroy Smith of the USACE. 'Some split due to hard driving. When compressing sheetpiling, we hit hard material at elevations of minus 58 or minus 60.' Smith says the piling started racking and fell out of alignment. The sheets, separated at the bottom, had to be pulled up, and Pittman had to start all over.

Ranjit Gujja, USACE project engineer for the London Avenue Closure, said the same strength of sheetpile was used when driving the second time around. 'We used a double template instead of a single template to give more balance and alignment.' Sheetpiling was driven between the two templates the second time around, to prevent racking.

The sill, a crucial element in the design of the canal, held the job up. On June 1, divers were cutting down the sheetpile on the west side of the sill, which was complete, and Pittman was in the process of completing the east side. The closure is expected to be complete by July 1. The cells were expected to be completed within a week. Discharge pipes were set, and the pumps were scheduled for a test run the next week."

[end update]

There's tons of other questions which must be asked, and which are not. This is just a sample.

Go to older posts Go to newer posts