Another way to clear the tubeside deposit is to bake it out in an oven. Many exchanger service companies have this capability. But this may not be an acceptable option if the turbulators are a non-metalic material.

Bake is an euphemism here, they actually burn it out. But hopefully in a gradual and controlled way.
There are a bunch of solvents that I have seen used, but it really depends on your coke.
There are consultants that work with nothing other than coke (Dr. Coke in Houston) and it sounds like you need to hire one of them.
OR, just rod or drill them out and trash the turbulators, then pressure blast clean.
The only real concern is will you damage tube.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

Turbulators should only be used in clean service (if you really must and not without specific Plant Owner approval), get the EPC contractor to foot the bill and for this and for lost production.

Hi Don1980, thanks for the info. So far nothing worked. One vendor offered heating the tubes with electrical heat tracing (wrapping the bundle)275Deg C and apply pulling forces on the turbulators. The tube sheets however, remain outside the heating area and are just covered with insulation. I feel 275 Deg C is not high enough but that is our normal operation temperature. I also would like to check with the manufacturer about the impact of the heating to the unheated tube sheets.

Hi EdStainless,

Do you have any link to Dr. Coke? Cant find it in google.
Thanks in advance

Looking at the pictures provided, are you sure you can't get a small water jet lance between the turbulator and wall of the tube? Do those turbulators spiral inside the tubes? They look like flat plates from the picture. If they are flat plates, you should be able to pressure wash each side, assuming the turbulator is centered and allows a cleaning lance inside. NLB and Jetstream both make tube cleaning nozzles that are 0.375" OD as catalog items. There may be smaller ODs available upon specialty request. I would contact your high pressure water cleaning company to see if this can be hydroblasted.

We used a low pressure water cleaning (5-6,000 psi) lance to de-coke our heat exchangers (not oil/gas application, but still had coking on HXs). I am not aware of any chemical cleaning solution that will dissolve, or even soften, coke. It seems the only options are mechanical cleaning via pigs (not an option here), water jet cutting (maybe?), and spalling via thermal shock (not great for tube/tubesheet longevity). The only chemical means of removing carbonaceous deposits would be, as others have said, reacting off with oxygen in an oven.

Hi,
I will try using the service of a third party contractor operating High pressure cleaning , let say 1200 bars unit!
Hope this is helping you.

Pierre

Hi TiCl4, unfortunately they are spiral. No way to enter. Once we are able to remove than we can use 15000psi to clean the tubes with a polishing nozzle. We have done it with our other shell & tube hx.

How about steaming the tubes with low pressure, high temp superheated steam? - coke would gasify to CO.

Difficult to achieve this we do not have high temp. superheated steam. We have decided to heat up the tubes in a PWHT furnace and pull the turbulators be pulling and twisting at the same time. Will update on the progress later.

In many countries, you should be able to rent a boiler. Drop say 100psi sat steam from the boiler to say 50psi and you have LP superheated steam. The higher the feed steam temp, the faster you can decoke these tubes without damaging the turbulators.

Thanks George, unfortunately we are on an island and there is no rental boiler available. Have you done such decoking before? It might work on tubes with fouling but blocked tubes with hard coke is a pain in the a.. We had to use manual removal with hydro jetting and lances.

Decoking furnace tubes with steam is a well known method to clean out coked up tubes. Agreed, you need to make a clear path through the tube to allow for a purge steam flow to occur. Which island in Indonesia is this plant in anyways, just out of curiosity?

It is in Batam, 20 km off Singapore. We produce TDAE as addituve for tyres and synthetic rubber.

Rental steam boiler offers from Singapore, just a short hop from where you are, up to 25barg:

https://www.atlascopco.com/en-sg/rental/products/s...

Update on turbulator removal..
The removal was extremely difficult due to availability and support here in Batam island. Covid19 did its part by limiting access of equipment and manpower. Anyway, we removed the bundles and heated the tubes up to 350 Deg C in a furnace (electr heated). It took total 2.5 weeks to remove most of the 1282 (2 x 641) pcs. The heating up had the following advantages but still it required other methods:
1. Heating reduced the viscosity of the residue (basically asphalt). The residue became low viscose and ran out of many tubes. However, the turbulators were like cemented into the tubes. Pulling was still impossible- twisted tapes were torn apart.
2. The heat exchanger maufacturer (not experienced in cleaning such tubes - and not the original manufacturer)used air lances to blow the tubes. This resulted in vaporizing the residue but also ignited the coke inside with the negative side effect of sparks flying out and therefore high temperature impact on the tube material. I was worried that we burn the tubes, which at the end did happen. We found 9 leaks and have to plug those. Still the removal was very difficult, we even had to use forklift to pull.
After 2.5 weeks we had all pieces removed with the result of currently 9 leaks and 16 tubes where rests of burned turbulators remain inside (melted and torn apart).
3. We did hydro jetting 15000 psi to remove coke and it works well.

So here is my current challenge, where I would like to get your feedback.
1. These 16 tubes with remaining parts of the melted turbulators might have damaged wall thickness (SS316L 25.4mm x 2 mm). We cannot meassure it as the tubes are not accessible. Hydro test 1.1 x design can only be done when installed in the shell as we do not have any test rings. This is no issue as long there are no new leaks coming up. It would require to remove the bundles again from the structure 30m high etc. Time consuming job under the current situation.
Question: Should we plug these 16 tubes too, where the turbulators were burnt to avoid possible leaks? I learnt that you have to drill a hole into the tubes, which are not leaking but have to be plugged. Drilling is also not possible. The reduced heating surface is not an issue for the capacity. But leaking thermal oil 275 Deg C in our product is something we must avoid.

I have uploaded some pics and videos if someone is interested to see the process of removal. It shows the heating furnace in operation, the pulling of turbulators, heat colours on the tubes, deformed tubes, hydro jetting and some damaged baffle plates due to quality issues during assembly of the heat exchangers)
Download link available until 04/08/20: https://we.tl/t-TqO4X52xvo

I would plug the (potentially) damaged tubes.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

Looks like there is some risk of rupture of the tubes when this HX is put back into service( and also reduced heat transfer duty performance), given the mechanical and thermal damage done to the tubes and turbulators. Company production management may be forced to get this damaged / compromised HX put back into service to reduce production downtime.
In the meanwhile, try to get a new HX:
a)Designed and procured with no turbulators
b)Generous tube side fouling factor
c)Low tube side dp in the unfouled design case, which will allow for dp creeping upward when the tubes coke up and still manage to keep hot oil flow at design rate
d)Which has a lower tube ID skin temperature when in operation
e)Much preferably a HX that can operate with lower hot oil feed temperature to reduce coking risk.
f)Mechanical cleaning of the tubeside should be made possible

This will surely be a HX which may most likely need a larger / longer footprint. And also find out what other operational and design improvements need to done to prevent this incident from recurring.

If there is a small tube leak when this damaged HX is returned to service, how will you detect it soon enough to prevent asphalt contamination with hot oil? Low level alarm in the hot oil expansion drum? Is there also a risk of reverse flow of asphalt into hot oil through a leaking tube during some operational upset?

Yes George, long therm and expensive solution. Turbulators in any process fluid that can cause fouling and can not be chemical cleaned is a no-go. Coke is just basically inert and diifficult to remove.

I know of an application with parallel HX and switch each month and then clean the one out of service. They can go about a year on one but buy then the unit is scrap. If they clean monthly they last 10 years.
Do you have room to go parallel?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

Yes Ed, but plants are designed for normal operation and not for in-house created problems. Our heat exchangers are already in parallel operation and there is no space anymore. We will buy spare bundles now as the existing ones are probably not suitable anymore for long term operayion.

From the photo I saw your exchangers are condemned. Meanwhile you can plug the damaged tubes and try to "live" with that situation until you purchase new exchangers. You need to make a root cause analyses, why this happenned. Is this an usual situation? Or something went wrong that the operators of the process doesn´t want to reveal? Ed gave you a good input "an application with parallel HX and switch each month and then clean the one out of service" you will stay with two trains, one spare of the other.With study and imagination you will discover more space. In house problems would be avoided following operating windows and procedures, the root cause analise shall be done.

good luck

luis

Hi Luis,

It all started with operational issues. Hx was heated but process fluid stopped so the solvent degraded and formed coke. Cleaning was not possible since the turbulator had to be removed which was the major issue and took 3 weeks. So we will no longer use turbulators as the design capacity is still enough. Right now both hx are running well but we will need to replace the bundles.