At the same time as I building my butane sphere I was also “Saving the Whales” or rather the dugongs.
You see, the BP refinery was sitting at the mouth of the Brisbane River on a small peninsular of reclaimed land from the 1960’s. Every 6 to 36 hours a huge crude oil ship would turn up to disgorge about 100,000 barrels of oil so this 80,000 barrel per day refinery could continue to operate.
Where the river enters the ocean is called the Moreton Bay Marine Park. The park is the only place in the world where significant populations of dugongs (and sea turtles) can still be found close to a major metropolitan centre. A failure and leak of the crude oil delivery pipeline would be devastating.
And guess what… The crude oil delivery pipeline had cracks in it. Yes, it was going to leak. Soon.
Snaking over the Brisbane River and supported on marine pylons about 5 meters above the water, this 20″ pipeline installed in 1965 was designed in the day for ambient temperature crude oils. Light, clean crude oils running at the temperature of the water it was suspended above. The allowance for thermal expansion (steel expands at about 11 x 10^(-6) meters per degree Celsius) was a series of direction changes placed in the pipeline’s length.
There where 4, simple butt mitred welds holding the pipe together. The quality of the welds where also terrible and not helping matters. The welds where like a seagull had had a dose of bad sardines. I am a welding inspector also.
There’s nothing wrong with a design like the pipeline had. Until that is, you start to run hotter crude oil through it. By the time I had arrived the refinery in 1994 – my first gig after leaving uni – the refineries’ inspection department was dutifully monitoring and reporting on the growing cracks in these mitred joints. The crude oil the refinery was processing was heavier and waxier and needed to be pumped at temperatures around 60 degrees. That’s enough for a bad skin burn. The pipe was flexing and moving so much it was pushing other service pipes off the shared pipe support.
And the main pipeline feeding the refinery was cracking.
I was tasked to fix it.
Quotes and estimates from “consultants” where coming in above AUD 10, even 25 million to make the repairs. And I did it for less than AUD 250,000/-.
This is what I did.
First I put aside the excitement of such a large project – I had only been at the refinery for 6 months, and had already saved several million dollars by the time this task was given to me (that’s another story). Then I walked the pipe. Up and down the 1300 or so meters above the river, breaking every HSE rule there was. And the 1,000 meters or so on dry land. I was looking at the way the pipe was moving. The cracks were clearly visible, made even more so by the white die being used for the mag- particle inspection. I’m also an NDT inspector.
The next thing I did was create a stress analysis model of the pipeline using AutoPIPE – a cool and pretty straight forward tool. I created the model from scratch using old refinery blue prints and verification on site. The model showed that the cyclic stresses where growing the cracks but they would never go critical and rupture. That was good news. But a leak is a leak and it had to be fixed.
By the way, if you don’t know much about metals, steel has a “critical crack length” which means that when a crack gets to a certain length in a piece of steel (or any metal or material too), the piling up of lattice dislocations locks up and the metal experiences a brittle, sudden failure. The 1 inch thick steel of the pipeline walls was below the critical crack length at the stresses it was experiencing so the metal would continue to tear and never rupture. It would leak into the river. Not explode into the river.
The next thing I did was walk the pipeline again and I noticed something interesting. The consultants were telling me that the only thing to do was put an expansion loop out over the water section of the pipeline. New marine pylons would be required and that is where the expense was coming from. I can still see the consultant rubbing his hands together now. [Protip: never ask a hairdresser “do I need a haircut?”]
But something didn’t feel right. There must be an easier way.
When looking closer at how the pipe was moving I dived into the pipe design code ASME B31.3** and studied the appendices. And there it was: a beautiful mathematical formulae showing stress as a function of bend pressure AND bend radius. The larger the bend radius the lower the pipe stress. I could dial in a crack free stress by adjusting the bend radius of the pipe.
In the refinery piping game there are either Long Radius bends or Short Radius bends. This is what ASME B31.3 covers and people rarely go into the detail to use anything else. But there wasn’t enough space for either of the standard bends using the existing marine pylons. They just didn’t work. So, jumping back into my AutoPIPE model I set a temperature limit of 85 degrees (way above any crude oil that the refinery was capable of processing) and worked out what the acceptable radii needed to be for each change in direction for the pipeline. There were 4 changes in direction out over the water that needed special attention.
Then, armed with my trusty tape measure I went back out onto the pipeline (thanks HSE) and checked if there was enough space for movement at my high temperature design with special long radius bends. There was.
Eureka!
The next step was to triple validate my calculations. I re-did them manually. I talked to everyone I could about my design. Even the refinery’s main advisory consultant – a former employee and friend of the refinery manager that found an office outside the fence paid more than one inside. He could find nothing wrong with my approach but he refused to validate my work in writing. I think I know why today. So I finally sent off a CYA letter to ASME in the USA, copy to the refinery manager, explaining that everything I was doing was within code and to come back to me if there was a problem.
They never did, and I wasn’t waiting for them to answer. I knew what I was doing.
The work would be hot, meaning there could be explosions if things went wrong, and that meant lots of precautions. The refinery manager had also made it clear to me: not a single drop of oil could reach the river. Of course. Was I stupid? Dugongs don’t like oily food. They eat sea grass.
Then in short succession I drew up the designs, had the 4 special bends made in Sydney and shipped the 900 km to the refinery, tendered and awarded the construction work and worked out a plan to undertake the repairs all within a 12 hour working window so there wouldn’t be a refinery shut down or a hold up to the next ship wanting to arrive and deliver oil. I had 8 crews working simultaneously to cold cut, bevel and prepare the pipe and insert the prepared new pipe spools. There would be 20 cold cuts being made at exactly the same time. It was pretty cool to watch, and Hans Walter, owner of the construction company doing the work was amazing. I learnt a lot from Hans and his leading man.
I had earlier negotiated with the operations team to hold a crude delivery ship over longer than necessary (demurrage is an expensive thing) and pump 100,000 barrels of sea water through the pipeline. I had another project on the go: raising the height of a crude oil storage tank by adding an extra strake so secondary seals could be fitted to reduce VOC releases (Tank 104. That’s another story). The tank was ready for hydro-testing, so I needed to fill it up with a lot of water fast. It was a perfect synergy: clean the pipe, fill the tank. The water would be released back into the river using the refineries biological water treatment facility. Oil is a natural material after all and bugs will eat it if given the right conditions. The ops team where more than happy to oblige. By the way, I held the next ship for a few extra hours too, getting it to fill the pipeline with water so we could hydrotest it – an important final task to ensure there where no leaks.
In preparation for the opening of the pipe I had a survey done of the pipeline level so I could calculate how much water was still laying in the pipe once it was washed out. Then I negotiated with the Port of Brisbane Authority to “borrow” several floating barges and had them fit them with tanks sized according to how much water would come out at each location. Later analysis showed that zero oil remained after the last crude ship had filled my tank with 100,000 barrels of sea water the night before. I couldn’t even make my hands dirty by rubbing the inside of the pipe when it was finally cut open in many places at once.
It all went off without a hitch. There where three notable events:
- During the draining of the remaining water (holes where hand drilled at the two of the main sections to be cut out – that’s what my pipeline remaining water level analysis showed) one of the drain hoses fell out from the tank on the floating water barge. I leapt down from the scaffolding and in a single bound – like superman leaping over a tall building – launched over the tank and sliding down the other side, in a single motion grabbing and re-inserting the itinerant hose back into the tank opening. It happened so fast no-one had even moved from their initial shocked position. The water than escaped stayed on the barge.
- The refinery manager criticised me for leaving “6” flanges over the water”. There are 10 of them. They where needed to pull out the welding balloons used to ensure a gas free safe welding environment. These 6″ nozzles had been attached and hot tapped in preparation for the 12 hour working window, as had most of the work, pipe supports, earth works on land etc. But I wasn’t worried. I had used special spiral wound gaskets with extra sealing capacity far beyond what would have been necessary. And written up a special procedure for closing the nozzles which including torquing the bolts, There’s never been a leak.
- I lost my temper once when at one of the bends, the smallest one, there was a lone worker who didn’t wait for the draining to be completed, and started his cutting too early. Luckily he was at a high point in the pipeline and luckily I arrived to check in on him as the water started to escape. He stopped cutting and held in the leak with his hand once I stopped yelling at him. It took a while before the water drained away from his location and he could continue his work.
The whole project took 5 months to complete. It was mostly a coordination exercise and it worked perfectly. It wouldn’t have happened without the full support of all the team players to carry out the strategy once it had been determine as sound and robust.
Full credit goes to my favourite Russia Alexy Lydov. A veteran at the refinery and in his late 70’s, he was there, direct from Russia, when it was built 1964/1965. I made a point of discussing with him everything I did during my time at the refinery. It was his idea to use welding balloons to enable the special joints to be welded in safely. That made the whole plan workable. His practical and clever Russian style thinking aligned with my get-it-done, can-do approach. He also knew where all the refinery blue prints where, and every design aspect that was relevant to know.
** I wanted a more rigorous design code than what you get from using API pipeline codes. It was a refinery after all
Recent Comments