Oil Spill report + TEDxOilSpill + Design proposal „Protei“ – TED Fellows

Posted in Uncategorized by matthiasdaues on Juli 12, 2010



26th of June 2010



3. Observations and difficulties for local sensing
These are the main spots we were able to sample oil contaminated water and oil accumulated on the containment booms. 


The water is very busy, so many boats, skimmers, converted shrimpers, barges, floating structures, static structures.  
Shrimper cleaner, Bay of Mexico, OIl Spill

Each black point you are seeing on this map below is NOT the trace of an oil spill. Each point is an oil platform. Read again, yes. Only one 1 of these point failed, creating the most environmental  trauma in the history of the continent. The Gulf of Mexico is „dangerous mess„. 

Image via <Chafic Kazoun>

And if you drive on land you will see the same thing : crude oil refineries everywhere
You will understand why Louisiana seemed to be more controlled by the oil industry than by the state itself. 30% of the USA needs in oil are covered in the Gulf of Mexico. About 50% of the US fishing industry happens in these fertile waters…
Oil Refinery. From New Orleans to Grand Isle, Louisiana

Back in Barataria Bay : it is very difficult to observe oil sheens or smell them (with our boat engine). Surface observation is difficult and expensive ; if we were to build a fleet of unmanned semi-autonomous drones, low cost local sensing is unlikely. We may rely on aerial observation (as the video below shows), high-altitude NASA reportsNOAA daily reportscrowd-sourced mapweather balloon shoreline reports and imagery mash-ups.
[archive .flv 39.7mb

We did not have to sail very long to find one of these.suspension

Accordng to an expert we met, this emulsion is typical of crude oil that has been treated with highly toxic oil dispersant Corexit EC9500A and EC9527A, produced by NALCO, a firm closely related to BP…

[archive .flv 8.4mb]

From experience, I can say that the contact with contaminated oil had no immediate effect on my skin. It is only 3 hours after I experienced very strong itching. At TEDxOilSpill, marine toxicologist Prof Susan Shaw from theMarine Environmental Institute confirmed effect of corexit to „dilate tissues and cause strong itching and perhaps skin damages“.P1030518

These are my hands

These are my legs … NOT ! I m much more hairy than that + I don’t varnish my nails (forgot one!) to visit a spill! Skin rash and blisters after wading and walking in Mobile Bay, Alabama, on May 11. [source]

One of the main problem of the dispersant is that is changes the consistency of oil, and the droplets end up everywhere in the water column as this little experiment below demonstrates.

[archive .flv 18.5mb]  

Crude oil changes properties constantly by evaporation, oxidation, emulsification, spreading, dissolution, biodegradation, dispersion, sedimentation.

It is very difficult to state how much of the oil is in surface on in the water column. In fact a large underwater plumes have been observed by Sam Champion and Phillpe Cousteau JR (video below). Prof Susan Shaw also dived in the spill and reported air pollution that „feels like my throat was on fire“. The spill is not only damaging the water, but also the already polluted air, Louisiana Bucket Brigade and the Louisiana Environmental ActionNetwork has been monitoring air degradation before the oil spill. Between 10 and 75% of crude oil mass evaporates in ambient temperature, so monitoring air quality may be of high relevance to understand the extend of the incident.

[archive .flv 35.8mb]

4. Containment & Sorbent booms, Wild life impact 


Many of the islands we sailed around are natural reserves, mostly bird sanctuaries. This is the worst time of the year it could happen : „love is in the air“ as Phillipe Cousteau JR said, this is the period of reproduction of many animals, including aquatic mammals like dolphins. Birds are the most spectacular victims of the spill, but with the extensive use of dispersants and the natural change of consistency of oil, a strong impact can be expected at the bottom, on the sea floor where it doest strongly damage invertebrates habitat. All the food web is affected, crabs, plankton, small, big fishes and ultimately humans.

[archive .flv 5.1mb]  

Even if many natural reserves are protected by containment (red or yellow) and sorbent booms (white – clean, brown – saturated with oil), they are very affected by the spill. Here, the low/high tide brought the spill high enough to cover the island.


This was the cleanest containment buoy we came across. Everything is sticky and smelly. 

containment buoy

Sorbent booms react very differently to different densities of oil : 

1- This buoy just below was exposed to very liquid oil, oil is absorbed all the way to the heart of the boom, the weight increased enormously. In average a 3 meters long sorbent boom, 13 cm in diameter, absorbs up to 113 liters of liquid oil (absorbing about 25 times its own weight), for a cost of ~100$. That’s 37 liters per meter, costing 33$ per meter. In large quantities the cost can surely drop. Important observation : the buoy saturated with liquid oil increase greatly in weight, at  the point that some buoys sink in the water (!!!).  


2- this other buoy (below) was exposed to very thick oil : only the surface is coated with thick oil, the heart of the buoy was still completely dry and white, unaffected. The buoy is very light. 

In order to make sure the both – buoys filled with liquid oil (heavy) | and | – buoys covered with thick oil (light) I put both samples to dry for 3 days. Neither of them dried, so we could confirm that it is oil, and not another aqueous solution.

Oil spilled sorbent boom
One of the main problem of the polypropylene sorbent buoys is that they are difficult to hold in place when clean and light, they fly around and get tangled. When they are saturated they dive and mostly disappear under the level of the water. P1030508

 Since the islands are inhabited by a large population of birds, a lot of organic materials fly around the island, branches, leaves, algae and feathers that are the best organic oil absorbent known (like human hair), all mixed with oil.Bay of Mexico, Oil Spill booms

When too close to the shore, buoys get washed on the rocks and get wrecked out. Wrecked bouys on the canal

Containment buoys and sorbent booms are useful, even if they are no silver bullet. It is not possible anymore to get close to buoys, which means it is making the sampling work and observation of many island very difficult. Medias said they will need to film boom-protected island „via telescope“ and complain about BP and government keeping them from accessing site and producing shocking images.Sorbent boom maintenance crew

These days, they are installing these huge barges anchored to the bottom with vertical trees. They pump the oil that runs across the surface in these massive trucks. This is only possible in shallow waters in moderate tropical storm conditions and there is a limited number of barges available, so it is great idea for this particular location, but doesn’t really scale up. I also saw immense metal booms, basically steel tubes floating around.P1030484


5. Fishermen 
As the sun comes down, oil skimmers are heading back home. Unmanned cleaning vehicles may not need to rest and could work continuously. I tried to speak to fishermen-cleaners and volunteers in Venice and Grand Isle but nobody would speak, too afraid to loose their jobs or expose their families. Nobody looked happy. FIshermen are generously paid by BP may they be collecting oil or not, many of them are  being paid even if they stay in the harbor I was said. There is such a strange relation between BP and the coast guards as well, why anyone would still want to hide the truth when it is all about working together the clean the Gulf? What is left to lie about?Cleaning vessel

You can see that shrimping boats are using the same combination of containment and sorbent booms. 

[archive .flv 4.1mb]

[archive .flv 33mb]  

6. Politics, Economic & Social
At TEDxOilSpill Latosha Brown underlined the affects of the oil spill is not only environmental but also directly on the people living in the area, energy trade, seafood, the culture of the region, tourism and national security. According to Brown „BP is acting like a superstate“ and another recent leak of information brought out the fact that many Vietnamese fishermen, representing up to 50% of fishermen in Lousiana, were being paid less than the half of white and black american fishermen, adding racial discrimination on top of this all.

I know little about the social and economical effect of the oil spill on the US economy, because I focus on design. Many argued it was BP’s fault, other pointed their fingers to the Obama administration, and some humbly declared that we were all collectively responsible for this, each of us being involved in oil consumption.

7. Health 
Similarly, the health effects of the oil spill differ greatly if you are reading BP, local hospital or NGOs reports. Thewikipedia article offers a satisfying range of the different viewpoints. As the demand in oil spill cleaners goes up, alarming articles state that most Alaskan oil spill Exxon Valdez clean up workers are now dead, with an average life expectancy of 51 years, dropping by 27 years US national life expectancy (78.4 years old). Still the cleaners have been required by BP not to wear respirators to „avoid spreading hysteria“. Corporation public image matters  much more than cleaners lives.

[archive .flv 6.4mb]  

[archive .flv 5.7mb

You dont need to work as BP cleaner to be affected, if you simply live in exposed area near the spill, or if you happen to be downwind on downstream, you will be affected. Many have reported that it has been raining oil inlands, farmers only start to report the effects of the oil spill on food culture.

[archive .flv 2.9mb

[archive .flv 5.3mb

8. Mapping and sampling
If cleaning oil is too dangerous, a least dangerous activity is mapping, air, water, sediment and waste sampling, or simply reporting via SMS, MMS, email, posting pictures, videos what you see, smell or feel (symptoms)…

[archive .flv 7.4mb]  

Grassroot mapping :
Crowd sourcing : 
Air Monitoring :
Water Sampling : 

There are so many initiatives down there, you can also donate money or equipment, so many things you can do locally or from a distance.

9. Bio-remediation 
There is a debate about „mother nature taking care of balancing this mess“. The process described at the TEDxOilSpill by Prof. Ronald Atlas called bioremediation showed that fertilizer application can hasten the rates of oil removal following oil spillages. He has worked on several major oil spills and explained that dramatic oil spill effects may be negligible in about 10 years. Local population victim of the Exxon Valdez oil spill contradict the „negligibleness“ Prof Atlas advocates 20 years after the spill. Also the Deepwater Horizon is happening at amuch larger scale than the Exxon Valdez.

[archive .flv 9.7mb]  

10. Hurricane season & Long Term Forecast 
The hurricane season has already started, making the oil containment effort nearly impossible in stormy days. In 2005, hurricane Katrina path was exactly where Deepwater Horizon oil spill origin is located. This year forecast predicts an intense tropical storm activity. Population evacuation plans are now set, trying to not forget anybody behind. This year has the worst weather forecast since 2005, remember Katrina (175 mph  – 280 km/h peaking winds!) :

[archive .flv 10.4mb]  


We can expect tropical storms to transport oil inland. we cant have humans risking their lives on the water during the storm, and If we make drones, they must be hurricane ready.

NOAA -shy- public announcement of the forecast for this year hurricane + oil spill interaction.

Helicopter view of New Orleans after Hurricane Katrina [source]

File:Thunder horse, oil platform, sinking, July 2005 U.S. Coast Guard picture.jpgView of Thunder horse, oil platform, owned 75% by BP, 25% by Exxon, sinking, July 2005 (before Hurricane Katrina, with a less violent hurricane), U.S. Coast Guard picture. 

As much as immediate hurricane response, NOAA long term predictions can help us design the response.On the other hand complex currents and winds make predictions very difficult, and the situation out of control, especially concerning less visible underwater plumes. 


[archive .flv 1mb]  

11. Sailing sorbent boom
11.1. Integrity

At MIT I tried to incorporate all these variables in the design of an oil recovery system. Unfortunately, in my department, the SENSEable city lab, my supervisors declared „it is too late to do anything for the on-going oil spill in the Gulf of Mexico“, and they decided to focus on „future technologies for future oil spill“ with very complex expensive robotics, absolutely wanting to include MIT developed oil-absorbing nano-fabric that may only be ready in 5 years for an estimated cost of 1000$ per square meter. 

I believe it is still time to develop technologies to clean up the oil spill. We must try.

I have resigned from the MIT SENSEable city lab 4 days ago. It was a hard decision to make because : If you have MIT students and staff brain power, technical facilities and world respect, there is no better place to invent and get supported to implement a solution for the Oil spill.

I also found out that an MIT student public call, by Nora O’Hickey, to foster MIT power to solve the problem was not even commented by MIT Director Hockfield and Prof. Slocum (who is both Obama and BP advisor, under non-disclosure agreement with BP). 

I am moving to the Gulf of Mexico next week. I want to work with people that also want to work for real. I believe I will be more useful there, either developing the following design, or, contributing to on-going researches and initiatives.

11.2. Deductive design

The deductive research process and design proposal I am presenting below has been denied by my ex-lab because it is not a „future technologies for future oil spill“, but an idea we may be able to implement as a product within a few weeks. I don’t know yet how well/bad it would work, we need to test.

11.3. Three Orders of priority of response

  • order 1 – cut at the source
  • order 2- capture the immediate output
  • order 3 – collect spill in the environment

[archive .flv 2.8mb]  

BP has already been sending many underwater vehicles to shut off the well valves, without success. Order 2 solution, capturing oil straight out of the gusher (ie. Top kill hat) are not entirely effective but they can significantly diminish the amount of crude before it reaches the surface. Both order 1 and 2 are achievable only if you have access to the origin of the spill (only BP), a fleet of underwater vehicles and an important budget, which is not the case for most, therefore I am focusing on order 3 solutions – collect the oil spill in the environment.

11.4. Functions

Robotic oil spill remediation

This is the chain of events necessary to make an autonomous oil collecting robot.

1. Sense oil
As described earlier in the post, from experience on site, visual sensing of oil is very unreliable since oil comes in so many different concentrations and visibilities, from almost invisible surface sheen to thick brown „paste“, to semi-solid tar balls. Video equipment (camera + pattern recognition to pilot the machine) is unreliable and expensive if the processing of the information was made on board. If the process of the image is made on land, after long distance video signal broadcast, the reliability and reactivity of the system would quite low and the cost very high. We use Fluorometer in ROV, but if you want to mass produce this machine, it might increase the price too much. If the machines are cheap enough to be produced in a great number we may rely on GPS : simply store the departure point and aim at one point in the center of the predicted spill, performing oil collecting maneuvers.

2. Collect oil
There are many ways to collect oil, but as time is pressing I choose to re-confgure the technology we already have and that works best. What everyone’s using both passively to surround and protect islands, as well as actively being pulled by oil skimmers is a combo of containment (yeallow or red) and sorbent boom (white when clean and light, brown when saturated with oil mostly under the surface). We can use the change of buoyancy of the boom (on top of water when empty – mostly under water surface when filled) to trigger a serie of switches that would say „empty“ / „full“ and send back the robot where it came from.

3.Extract oil
In the future we can think of continuous extraction (watch video), but that would mean the machine is equipped of a „washing machine“ that extracts oil and stores it in several tankers after processing without affecting how well the boat wouldsails (balance).  

4. Process (crude) oil
Processing crude oil is the refinery process. Just as the previous step, ideally this process happens as continuous distillation. Refining crude oil is a complex and dangerous process, that requires monitoring and present high risks of explosion, high maintenance, elaborate machinery, high cost. As far as I know, mass-producing mini-mobile refineries would be very expensive and could potentially transform every oil-collecting-robot into a mini oil spill of its own.

5. Re-purpose oil
If a unit is capable of sensing, collecting, extracting and processing oil, it would be producing combustible (stored energy) on one hand and raw material (like asphalt) in the other hand. You don’t want to use the oil you collect and process as main combustible to move your machine around, because for long periods your machine may not be exposed to enough oil to propel itself. Another propulsion system, like a sail, uses what is almost available and free all the time, wind. Similarly you don’t want to accumulate and carry around asphalt or highly toxic materials.

MIT SENSEable city lab fro wants to make a autonomous robot that makes the 5 steps into one machine, so it is a very long term plan in the context of low cost swarm (many) robotics. I want to implement the technology now, because we need it now.

Locally sensing oil is expensive and kind of unnecessary if you have good predictions of oil spill (esc. step 1). You would only want to extract oil on board of the machine if you want to process it (refinery) but this is a very complex, dangerous and expensive process (esc. step 3,4,5). The quantity of oil that needs collecting is so immense that we need a technical solution that is simple and would scale up easily, so we can only focus on event 2. : collect oil. 

11.5. Requirements

  • Collect a lot of oil, bring to a safe collector : ultimate goal, collected on land or to a large „mothership“.
  •  Unmanned : to avoid exposing human cleaners to toxic oil and gases. To operate day and night without interruption. To operate in hurricane conditions.
  • Self righting : boats do tip over. Only buoys, which have low maneuverability passive structures don’t tip over.
  •  Self steering : the machine needs to have high maneuverability, especially the capacity to go upwind, where the oil is coming from and stretches in long sheens.
  • Unbreakable : hurricane-ready design. Doesn’t mean titanium strong, but perhaps lightweight and flexible
  •  Autonomous : if we have electro-mechanic and electronic on board, the machine should have comfortable autonomy probably provided by solar panels and batteries.
  • Cheap : we have to protect hundreds of miles of coast, more than 6’500 km2 of spill, the machine needs to be simple, robust , reliable, cheap and easy to manufacture in order to scale up and deploy.

11.6. Intercept oil plumes
Current oil skimmers cut a clean line through a sea of oil. Even A whale super skimmer do not offer a large capture area while consuming vast amount of fuel to move around pulling booms.

[source of the 2 above images]

The bigger picture is that in most case long oil plumes are going downwind, with superficial surface currents and waves. Therefore the best way to capture an oil plume is to go upwind. We must use the forces available. Use the force of the hurricane to capture the oil spill. If we go upwind and tack left and right, we may oppose the natural flow (near perpendicular interaction) and collect a fair amount of oil in the successive folds. 

Oil Collector swarm boat

Now if we have many of these units we can work on the behavior and trajectories that allow more oil interception.
Oil Collector swarm boat

Sea Swarm, collecting the oil spill

11.7. Sailing upwind pulling a long sorbent tail
So, in the end, it is about engineering a „propulsion head“ to pull a long line of sorbent boom. Only a hydrodynamic profile could efficiently sail upwind. 

SeaSwarm exotic colors


11.8. Steering
If you are pulling a long tail, you are adding significant drag to your structure. A classical boat has an articulated rudder at the back to steer, and a center board at the center that acts like a pivot. Since we want to add the longest tail possible, steering at the back gets increasingly inefficient, so the idea here is steer at the front, not with a rudder but by changing the general geometry of the boat. It works as a series of vertebrae and required only a cable to steer.

Oil Collector swarm boat

Penguins (swimming vertebrate) and the fabulous AirPeguin by Festo uses this principles in 3 dimensions (also up and down):

[archive .flv 9.9mb]

11.9 Cheap inflatables, unbreakable, weight distribution
Inflatables are cheap, easy to mass produce, safe (they rarely cause damage in collisions), lightweight and of course float very well. In order to distribute mass and adjust buoyancy inflating with :

  • air : for the sail
  • water : for the hull
  • sand :  for the ballast

may provide a very reliable, cheap and safe design. 

Oil Collector swarm boat

11.10. Self-righting
In a hurricane context (winds peaking at 280km/h) it is vain to oppose wind. It is preferable to tip on the side momentarily, the sail would naturally „stick“ to the water, meanwhile having enough ballast at the bottom to have the boat self-right to navigate again when there is bearable wind. A passive and flexible system.

Oil Collector swarm boat

11.11. Electro-mechanics
Steering such machine would only requires 2 winch step-motors : one to bend the „nose“ of the boat left or right ; another one to sheet the main sail. We could add to the machine the high visibility flashlight at the top of the mast, the GPS receiver, an accelerometer, a circuit board for general control and to store the start point and destination coordinates. All electronics may be sealed dry near the mast by the center  of the „boat“. A large battery, recharged on land and solar panel maintained may provide energy for the electro-mechanics.


11.12. Testing 
I made a quick and dirty test machine that was quite promising. 



We (with David & Eugene Lee + Sey Min -> Korea Power team!) added to this test model a lot of stones at the bottom for ballast to sail in the Charles river in Cambridge MA. The mast was not only inflated, it was re-enforced with plastic plumbery tube.



test of oil spill collecting drone 20100623


Next steps : pulling a long tail and testing oil collecting trajectories + more extensive test of front steering.

12. Open source technology and manufacturing

There is still a lot of collaborative research to be done on the whole project „Protei“ to see if it is a competitive proposition to collect oil, but in order to insure the future of the research. 

Creative Commons License

„Protei“ by Open_Sailing is licensed under a Creative Commons Attribution 3.0 Unported License.
Based on a work at Permissions beyond the scope of this license may be available at

This does mean that everyone is invited to use, make commercial use of the technology, but they must aknowledge they are building on top of our open-source technology licensing and themselves share their designs and advancements. This methods allows collaboration as well as commercial development of the technology, because we will definitely need more than one company to produce this, and a lot of caring clever people to develop and fork the technology.

You can follow and contribute to the development of Protei on Hosted by Open_Sailing. Feel free to contact me <contact [at] cesarharada [dot] com> for more informations and/or to get involved.

Because corporations are so good at re-writing history, I invite you to download all the videos on this post for your archive in a zip file [209mb] and be ready to re-post them if they „disappear“. 

13. What’s next?
As I said earlier, I resigned from MIT SENSEable city lab in Boston to move to the Gulf of Mexico where I believe I will be more useful to either develop this technology or contribute to existing projects. I am now looking for :

  • partner experts, university, inventors, sailors, RC boat hobbyist groups
  • partner Industries : inflatables, sorbent manufacturers, marine robotics
  • a sail boat to test the idea and take the prototype to the oil spill
  • a workshop around New Orleans to fabricate more test models, ideally close to the water
  • contacts with local authorities to get permission to navigate and test prototypes
  • safety equipment (Hazmat + respirators) for our crew 

Thank you very much for reading this long article, I hope it helped understanding the extend of the environmental challenge we are facing, and re-gain enthusiasm in the idea that we can solve this problem together. 

Posted via email from matthias daues‘ posterous


1 comment

Eine Antwort

Subscribe to comments with RSS.

Kommentar verfassen

Trage deine Daten unten ein oder klicke ein Icon um dich einzuloggen:

Du kommentierst mit Deinem Abmelden /  Ändern )

Google+ Foto

Du kommentierst mit Deinem Google+-Konto. Abmelden /  Ändern )


Du kommentierst mit Deinem Twitter-Konto. Abmelden /  Ändern )


Du kommentierst mit Deinem Facebook-Konto. Abmelden /  Ändern )


Verbinde mit %s

%d Bloggern gefällt das: