Heat exchanger (cooler) for high viscose aromatic fluid

[Stefan2211]

Dear All,

We have a by-product in our TDAE plant, which needs to be cooled down from 180 Deg C (7.4 cP) to 115 Deg C (107 cP) by means of water. The flow rate is 11,000 kg/h. Our EPC contractor provided a welded plate heat exchanger with 27 m3 heating surface. This plate hx was basically damaged beyond repair during an operational incident, where carbonized solids from process were circulated, which lead to blockage and shut down. The fluid became solid and could not be liquified anymore by means of steam (too many coke solids in the plates). Now we are thinking of replacing the plate type with a shell and tube single pass hx in vertical design (fluid in tubes - water in shell). What would be the recommended min tube diameter for easy cleaning?
In other section where we heat up the fluids we use 25.4 x 2 mm tubes in either SS or carbon steel but mech. cleaning besides hydro jetting (15,000 PSI) is not practical (manual brushing).
Does anyone have experience with such asphalt like fluids? We are thinking of using 42.2 x 3.56 mm (DN32) for easier use of rotating brushes etc.
Thanks in advance

 

[EdStainless]

There are many vertical HX (fall film evap for example) that are built with 1.5" (38.1mm) up through 2" (50.8mm) tubes. By being verticle you can get away with thinner walls that in horizontal. You have to watch flow distribution in the tubes though. It requires some sort of dam or restriction at the top of the tubes in order to keep product in all of them, and the tubes won't really be full (in your case you would want the tube walls coated at the top and the tubes just full at the cold bottom).

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

 

[Latexman]

We use 1" tubes in our S&T heat exchangers for easy hydro jetting.

Good Luck,
Latexman

 

[TiCl4]

Most high pressure manufacturers (Jetstream, NLB, etc) make tube cleaning nozzles as small as 0.375" OD for nozzle pressures up to 40,000 psi. Our most recent heat exchanger specification had 5/8" OD tubes for this reason - we checked with several of our hydroblasting vendors, and they indicated that 0.5" and up was good for cleaning for them (i.e. they had nozzles on-hand for the job and would not need to order anything special).

Also, I suggest you consider an TEMA "A" head for the top head if you intend to lance the heat exchanger in place. Instead of needing to lift the entire head off, you'd just need to remove the top flange. This sharply reduces the lifting requirements (overhead space, total weight) for CIP. This obviously needs to meet your other process needs, but it does present the easiest access for cleaning.

Also, don't discount multiple passes if you need to get your Reynold's number up (you mentioned a "high" viscosity fluid). As long as the passes are created by partitioning the head (not U-bend tubes), then you can retain the ability to lance the tubes even with a multi-pass design.

I've dealt with vertical S&T heat exchangers that had coking that needed to be cleaned regularly (every 5-7 weeks). Hydroblasting was the only effective method I saw used with that. I have no experience in how well a brush cleaning would actually clean the tubes.

 

[georgeverghese]

How did the coke solids get into the all welded compact HE ? Usually, there would be a guard basket strainer upstream to retain solids that may be entrained in the feed. If you wish higher availability, you can install 2x100% 40mesh basket strainers, piped up in parallel? Add a dp alarm in DCS to tell operators to switch over.
An all welded compact HE is one way of getting high Nre and reasonable htc in this application with high viscosity process side fluid. The shell and tube equivalent may occupy a much larger footprint due to the low process side htc.

 

[Stefan2211]

Thanks guys, so it seems that going for larger tube is not really necessary by yr experiences.
George: The plant design was based on clean fluids so only pumps were protected with 40 mesh. However, an incident lead to coking of our dilutant (overheating)and the coke built up in pipes , vessels and columns. Under design conditions there should not be any solids. Plate type hx can't be cleaned by hydro jetting due to the corrugated plate design.

 

[Stefan2211]

TiCl4: have you ever tried steamin out coke like georgeverghese has recommended in a previous thread months ago?
"How about steaming the tubes with low pressure, high temp superheated steam? - coke would gasify to CO"
George: Di you do it before? Is it a matter of a view days to steam it out or a long time during a real shut down?

 

[TiCl4]

Stefan2211,

No, we did not try that, as far as I know. Our application was a bit different than yours though; the coke was intermixed with oxychloride species, which are quite water reactive. Using high pressure water cleaning both physically removed coke and hydrolyzed the oxychlorides, losing the coke from the tube walls. An ~1,000 sqft heat exchanger would only take 4-8 hours to lance. Given that, I can't imagine steam cleaning would be faster than lancing.

 

[georgeverghese]

In my experience, all compact welded HE have guard strainers on all streams going into the HE, even in very clean service. Looks like a guard strainer would have saved your HE in this incident.

It is standard practice in refineries to steam out coked up furnace tubes if coke buildup becomes unacceptable and furnace performance drops. I've not done this myself. Also done routinely in naphtha and ethane cracker furnaces for C2= production.

 

[TiCl4]

Georgeverghese, what benefit does streaming out have over high pressure cleaning? More cost effective, or able to clean geometries a lance cannot?

 

[georgeverghese]

From what I've read so far, no particular advantages for steaming out tubes over hydrojetting.

 

[Latexman]

I had a case where steaming out was way better. I used to run a para-tertiary butylphenol unit. Phenol + isobutylene = PTBP. Phenol freezes at 41 C and PTBP freezes at 98 C. When the insulation and/or heat tracing malfunctioned, things started to freeze up and plug. As such, over the years we could just about steam out anywhere. Blast it with steam for a while, shut the steam off, and keep running! Then, fix the problem, and keep running!

Good Luck,
Latexman

 

[EdStainless]

When the coke is combined with something sticky then the heat from steam can soften and break it down, You have to be aware of the nature. With steam though you have to be prepared for the temperature excursions.

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

 

[Stefan2211]

We had sticky material due mixed with coke from overheating the dilutant. There were a combination of problems.
1. Using steam did vaporize the dilutant and caused hardening the coke. Basically it blocked some the tubes. Took a long time to use hydro jetting 15000psi to "drill" through it and flush it out.
2.The sticky stuff could be removed with chemical cleaning (as long there was a flow through the tubes) and the harden coke was only possible to remove by hydro jetting.

 

[Compositepro]

There are many unique designs of equipment for high-viscosity heat transfer. Screw extruders, rotating drums with scrapers, tubes with wall scrapers.

 

[georgeverghese]

From memory, another type of all welded compact HE is the spiral HX, good for high viscosity applications. Guard strainers may still be required-??

https://www.nexson-group.com/products/spiral-plate-heat-exchanger/spiral-type-1/

 

[Paul Nuttall]

Certainly the Spiral Heat Exchanger is a good solution for a fouling service. In this case it would seem that the fouling occurred by process upset in a usually clean service so you may still want to consider the plate exchanger with adequate protection. Moreover, even though the Welded Plate Heat Exchanger has a corrugated plate design it should still be cleanable using high pressure water lance depending on the specific type of exchanger that is installed.

 

[TiCl4]

Just to add a tidbit I didn’t see mentioned - spiral heat exchangers can be made to have the front plate removed, which allows full access to both hot/cold sides of the spiral for mechanical cleaning.