Four who died were not trained for Delta P - Trinidad and Tobago

[Endler's]

 

[ChrisDee]

I'm a little curious why OSHA is involved. OSHA is an USA regulatory agency, whereas occurred in another tiny country.

 

[tridacna]

even considering it's a tiny and poor country.

Why do you think that T&T is a poor country?

 

[Papyone]

Hi, to help you to have a better view of the real facts, I’ll give you here below a more precise information on the aftermath of this unfortunate accident.
Sorry, as the comment is too long to be published in one single message I’ve split it in 3 parts.

A bit more than one year ago, five divers were sucked into a pipeline. One of the divers managed to get out of the pipe on his own, but unfortunately the other four died on the spot after a few hours because nothing was done to save them in time.

Public hearings into the investigation of this incident that happened in Trinidad and Tobago on February 25, 2022 are now closed and the final report was expected to be released in April, but unfortunately it has now be postponed until November of this year.

But thanks to the hearings as well as the numerous documents that were published on this occasion, we can already realize that the main cause of this dramatic incident is due to ignorance by all the parties concerned (diving company/ divers/ customer) of this type of danger.

Nevertheless, in this investigation, a question remains unanswered as to whether these four divers could have been saved or not. It is of course not easy to answer this question, but by analysing the way the incident was handled after the return of the only survivor, the answer is definitely NO, and this for the simple and good reason that at no time did the people involved in this rescue worry about the depth of the water and the absolute pressure prevailing inside the pipeline as well as the time that was passing since the beginning of the incident.

Reminder of the course of events:

14h40: Four divers are working in a hyperbaric chamber placed over a 30” riser. The absolute pressure into the chamber is 1.45 bars.
14h43: A fifth diver arrives at the bottom of the chamber to bring a wrench to the team.
14h45: One of the divers deflates the plug that is present in the riser and the delta P occurs.
The 5 divers are violently sucked into the riser/pipeline.
14h46: End of suction.
14h5?: The rescue diver enters the hyperbaric chamber and finds that it is empty.
At that moment he also notes that the water height in the riser is equal to the lower level of the chamber (i.e. 4.5 m or 14,76 ft.).
The alert is given.
16h00: Bangs are heard in the pipeline.
16h25: One of the divers C.B. has managed to come back to the surface in the riser but he cannot get out of the water. He taps and shouts to be heard.
16h45: Two rescue divers enter the hyperbaric chamber and help Christopher B to come out of the riser. The latter asks them if he has any decompression stops to make.

Without bothering to check the inner depth of the riser, the supervisor replied no, when in reality he had just spent between 100 to 105 minutes at 15.7 m (51,6 ft.) which in the 55 ft table of the USN Manual rev 7 gives an air stop of 17 to 34 min at 6 m (20 ft). The fact that Christopher did not get bend is probably due to the fact that he spent several long minutes in the riser at the pressure of 1.45 bars.

16h48: Christopher is recovered on the surface where he informs the team of the situation in the pipeline, emphasizing the fact that his 4 colleagues are injured but alive.
He describes that three of them are in a large air pocket and the fourth is waiting in a second, smaller air pocket closer to the riser elbow as he has been following C.B for some time.
17h??: Following the information given by the survivor, the surface team prepares the hookah.
17h45: Diver M.K assisted by divers C.C & R.R enters the riser and then progresses in the pipeline to the end of its hose, i.e. over a distance of about 3 m (10 ft.). As he encounters nothing, he comes back up.
18h??: After his return to the surface, the team puts in place a new dive plan which includes sending a tethered scuba diver in the pipeline with a diving bottle equipped with 2 regulators (at this time there is no commercial diving equipment available yet).
18h25: C.B assisted by M.K & R.R are back in the habitat and are ready to dive, but at the last moment the diver feels that his tender M.K was not in a condition to assist him properly, so he decides to abort the dive and wait for commercial diving equipment to arrive.
18h30: Another commercial diver (C.B's brother) arrives on site with his commercial gear and a 91 m (300 ft.) umbilical.

Another dive plan is redone in which it is planned to send a commercial diver down the pipe until he reaches a missing diver. He would then tie a rope to him and let the chamber attendants pull him. It is not specified how the injured diver was going to be supplied with air.
18h30: Bangs are heard near riser B5.
19h00: The client decrees a total ban on diving despite the arrival between 19h00 and 20h00 of two dive vessels with full commercial diving equipment. The reason given is that the client and the Incident Command Team (ICT) believe that another delta P could still occurs and so they want to first inspect the integrity of the pipeline with a ROV or a crawler before allowing divers in again.
20h00: Banging is again heard near riser B5.
20h30 – 22h00: Upon request from the dive company, the command team (ICT) authorizes divers to install the B6 upper riser extension to prevent possible filling of the riser and chamber in the event of a failure of the compressor.
22h10: A company is contacted to provide a crawler.
23h00: Removal of the blind flange from the habitat to allow access to the upper part of the B6 riser which is now out of the water. The removal of this cover has the effect of reducing the absolute pressure in the pipeline by +/- 0.2 bars, but also has the effect of no longer allowing the decompression stop to be made in the hyperbaric chamber.
02h30: (Saturday) bangs are heard for the last time near riser B5.
03h00: First intervention of the crawler from the B6 riser side. As it progresses, the crawler encounters a scuba tank at a distance of 50.6 m (166 ft.) from the top of the riser. The craft tries to push the bottle away, but ultimately can only move it over a distance of (5,8 m (19 ft.).
05h00: After several unsuccessful attempts and seeing that the crawler cannot advance further, the ICT decides to send the machine to the other side of the pipeline and authorizes the diving company to remove the B5 riser blind flange.
06h00: The blind flange of riser B5 is removed. Nobody seems to realize it, but the consequence of this act has the effect of lowering the water level by about 11 m in B6 riser and thus at the same time displace this mass of water in the pipeline, but also and above all to completely and quickly drop the pressure that reigned in the pipeline (+/- 2.3 bars) to atmospheric pressure.

Decompressing these poor four divers who had been confined to this pressure for about 14 hours, in less than an hour sealed their fate and from there nothing more could have been done to save them alive.

(see next comment).

 

[Papyone]

Continued comment 1

Question:
By removing this flange, did the ICT and the diving company know what they were doing and what the consequences would be?
Answer:
Apparently not, because even a few hours later the ICT was still asking the dive company to draw up a new rescue dive plan.
Question:
Was the ICT correct by saying that a second delta P could still occur if the plug was not fully deflated?
Answer:
No

Here, in this incident, the delta P was related to the deflation of the obturator and occurred as soon as the pressure inside it became insufficient to hold it in place. Today we know that the plug open/close valve was torn off. This could have happened when the diver opened the said valve with his spanner or more likely because the valve got caught in the retaining chain at the start of the plug movement inside the riser.

In any case, in the real situation, it can be said that, as it moved through the pipeline, the plug quickly and completely emptied of its air due to the increasing pressure exerted downstream of it.

This means that at the end of the event which, let us remember, lasted about forty seconds, this plug returned to its original shape, that is to say that of a cylinder of about 15 '' of diameter.

Obviously, at the time of the incident, no one knew the exact position of the open/close valve and therefore one could also imagine (although this is more than unlikely) that the diver just had time to close it when he realized that something was wrong.

This means that in this case, the plug would then have moved in the pipe without the air being able to escape. Could this have resulted in blocking it again? The answer is off course NO because even if the bag was still practically inflated when it started to slip, it would then have been subjected to a higher pressure (2.5 bars) than it had at the start (1.45 bars) and this would have resulted in a strong compression and deformation of its outer envelope and thus a reduction of its volume and its initial diameter. This means that in both cases (open valve or closed valve) the water could easily have circulated between the obturator and the wall of the pipe and therefore could not generate another delta P. Why the ICT (which also included a dive expert) has not come to the same conclusion remains a mystery, but it has resulted in many hours being lost in unnecessary dithering.

But let’s assume now that ICT has not banned rescue dives. Would that have changed the course of events?

Unfortunately, until 19h00 the potential rescue divers didn’t have the adequate commercial diving equipment.

We saw that after the hookah dive which did not yield anything, a second rescue dive, planned to send a scuba diver inside the pipeline with a diving bottle (12 l) equipped with 2 regulators (it is not said, but we can suppose the diver had also planned to take another diving bottle with him).

Under such stressful circumstances and due to the difficult progression in the pipeline (feet first travel / portion of pipeline partly without water / wall full of slippery crude oil), the rescue diver would have consumed much more air than in a normal situation which means that with its 2400 litres of air he would have had an autonomy of about 20 minutes, and that during this period of time he should have, if it had found someone, correctly tie the retrieval rope, then give (or place) a diving mask on the head and finally secure a scuba tank properly on his back (or on his stomach depending on the injuries) while making sure it would stay in position during the ascent of the diver in riser, and to finish, he had to guide and return to the riser with the diver.

Knowing these parameters, we can say that with a scuba equipment, this rescue would have been very RISKY if the injured diver was found more than +/- 45 m (150 ft.) from the elbow and the risk would have been even higher, as nothing had been planned with the assistants on the dive time and the maximum penetration length in the pipeline.

As consequence, it can be safely said that it was a VERY good thing to have aborted this dive.

Subsequently, another dive could have been made around 19h00 with this time a 300-foot umbilical. This length would have allowed a penetration of approximately 200 feet after the elbow and would likely have been sufficient to reach the diver which was following Christopher (provided the diver remained in place). The problem here is that once the diver was brought back into the chamber, he should have, after his long dive (> 255 min) at 55 ft. stay there for very long air decompression stop (313–432 minutes) or (83-106) if O² was available, with the high possibility to face a severe decompression accident as this stop would have been made at a depth less than the depth of the table.

With the 55 feet table the maximum exposure time is 360 min (6 hours), and this duration was reached at 20h45. This means that from that time, the rescue of the divers would then have become more and more problematic because there was then no longer a decompression table available for such a dive time and therefore the risk of generating a severe or even more serious type 2 accident would have been inevitable.

(see next comment)

 

[Papyone]

Continued comment 2

It can therefore be said that, even if the dives had been authorized, the chances of success would have been extremely slim to recover more than one living diver, and even more if the upper extension of the B6 riser had also been installed since from that moment on, no decompression except decompression in the riser would have been possible.

Does this mean that whatever was done that day, these four unfortunate divers had very little chance of making it out alive? The answer is unfortunately NO.

What was fatal to the four divers was that as time passed they were increasingly saturated with inert gas (nitrogen) and after 20h45 any rescue attempts would more than likely have been doomed to failure.

Yet it would have taken very little for them, too, to more than likely be recovered alive from that pipeline.

What should have been done as soon as the only survivor returned, would have been to immediately look at the actual depth at which the other divers were confined because that would then have shown that there was going to be a decompression problem very quickly and that to avoid this, it was necessary to reduce the ambient pressure in the pipeline and in the air pockets as quickly as possible. And so the only way to do that without completely flooding the line was to lower the water level in riser B6.

This operation could easily have been done by lowering a submerged pump or even an airlift (at the start of the pumping phase) into the riser via the hyperbaric chamber and thus pumping out a certain amount of water (+/- 30 m³) so as to lower the water level in the riser until only about 3 m (10 ft.) of water remained in it, which would then have reduced the absolute pressure in the pipeline to (+ /_1.45 + 0.3) = 1.75 bar.

Doing so would not only have allowed the divers to already desaturate their tissues from the maximum pressure, but it would also have had the effect of increasing the volume of air in the pipeline.

As we can imagine, these rescue dives would not have been the easiest to do and before they could be done a certain number of points should have been adjusted in order to dive safely.

Among these were:

Launching and raising of divers:

Since a great part of the riser would have been drained out of its water, it would have been essential to install a lifting system in the chamber. This could have been done after the start of pumping during the waiting period for professional equipment.

Progress in the pipeline and travel distance:

After the incident, no concrete element could say at what distance from the entrance to the riser the divers were. It was therefore essential that the umbilical’s had a sufficient length to at least exceed the central point of the pipeline by a few meters. After the arrival of rescue boats, several umbilical’s of 91 m (300 feet) were available and they could easily have been connected to each other in order to have sufficient length. (To save time, these transformations could have already been done on the way to the site).

Another difficulty that should have been faced because of the limited space inside the pipe was the use of the bailout bottle. It would of course have been absolutely inconceivable to deprive the divers of this emergency gas supply, but in this case could it not have been replaced by a smaller bottle such as a pony bottle of 5 or 8 litres. The answer is obviously NO, because an 8 litre at 200 bar (1600 l) would only have provided autonomy of about 10 minutes which would have been insufficient to return from the middle of the pipeline.

Indeed, it should be remembered that the only survivor remained approximately 105 minutes in the pipeline. We can assume that just after the incident, a lot of time was lost by the divers to recover from their emotions and to decide to move, but we can decently think that the only surviving diver then took between 30 and 60 minutes to return to the top of the riser. This means that in the event of an incident occurring during the rescue dive, the capacity of a conventional bailout would also have been insufficient because it would only have offered autonomy of approximately 16 minutes.

Therefore, for this rescue, it would have been preferable to leave the bailout bottle on the surface while providing the air to the band mask or helmet via a second hose taped to the main umbilical. This technique is frequently used in inshore diving when working in very confined areas where evolution is difficult. The advantage of this technique is that it increases the autonomy of the emergency air because the bailout bottle can then be replaced by a larger one.

On the other hand, it is obvious that this technique can only be implemented on condition that the risk analysis demonstrates that during the dive, nothing can fall on the umbilical or hinder it to the point of blocking it and that the only way out for the diver would then be to cut his line. In this incident, with the exception of the 4 scuba tanks (which should have been brought to the surface before progressing further forward), and the plug (which must have been ahead of the divers), no other large obstructions was in the pipeline and therefore the safety conditions were met to leave the emergency gas on the surface.

As one can imagine, these rescue dives would not have been easy because to facilitate the return or the recovery of the rescue diver in case of problem, it would have been necessary to progress with the feet first over probably a fairly long distance (+/- 200 m) in a slippery pipeline including also partially filled sections with water which would therefore contain a relatively toxic atmosphere to breathe. At first glance, a logical choice would thus have been to use a diving helmet because it is more waterproof than a 18/28 band mask (MOD-1 not available on the site). But given the type of progression (partially dry environment) it would seem that a band mask would have been preferable because it would have been much lighter and comfortable to move around.

With this band mask, the risk of inhaling toxic gas fumes in the waterless sections would have been higher than with a helmet, but to avoid this, it would then have been sufficient to leave the free flow slightly open in such a way to have a continuous flow of fresh air under slight overpressure, as is the case with a hazmat helmet or mask.

Eventually, the use of continuous free flow would have had the effect of once again increasing the ambient pressure inside the pipeline and therefore also raising the water level in riser B6. Hence the needs to monitor and pump back the excess water in order to stay below the critical level.

On site, no dry diving suit was apparently available. Therefore, to limit as much as possible the contact of the wetsuit and therefore the skin with crude oil, the rescue diver(s) would have had to wear protective clothing (Tyvek or overalls) and more or less waterproof gloves.

Finally, and contrary to what has been said by certain diving experts, in NO case could the umbilical have gotten stuck in the roundness of the riser elbow and therefore the immersion of a second diver at the bottom of the riser was ABSOLUTELY not recommended because in the event of a problem it would simply have been necessary to retrieve the diver by pulling on the umbilical rather than sending another rescue diver into the pipeline.

If these few suggestions had been followed, and if no divers had been found beyond 200 m, (656 ft.) then the upper extension of the B6 riser could have been installed. This would then have allowed the opening of the B5 riser without any risk of decompression sickness and would then also have permitted to do the rescue from that side.

Unfortunately, we see that following the accumulation of errors, things did not end in the right way.

 

[Seameadow]

Thank you for the detailed report and feedback. No doubt very difficult to rehash all of this.

 

[CoralSeaMan]

Thank you for your summary.

 

[Sunn]

Papyone, this is a very good report.

 

[Papyone]

Papyone, this is a very good report.

Thank you Sunn, as you may have realized, even if the incident command team (ICT) had reacted appropriately, the rescue of these divers would not have been an easy task.