Crude Flashing ok in S&T? 1

[Rosalynn]

I have a project to increase the throughput in an asphalt plant. One of my ideas is to increase the temperature entering the first flash vessel in the unit, by making the feed/preheat train more efficient. There is no fired heater upstream of the first flash vessel and to keep op costs and cap costs low, I'd like to avoid installing one.

I am aware that flashing in heat exchangers reduces the effective heat transfer coefficient because of the differences in thermal properties between vapour and liquid, but are there other reasons out there for avoiding flashing in a heat exchanger?

I have a backpressure controller on the outlet of the preheat train, so I can suppress flashing in the exchangers to a certain extent; however, the supply pressure is limited. I'm looking at up to 13 mol% flashing, but expect the economic option would be somewhere between 5 and 10 mol%.

 

[Montemayor]

Rosalynn:

Your scope of work is expressed in pieces and is very confusing. This is not a complaint or criticism; this is a request for clearer information so as to make out what you are trying to do and in what manner. In the absence of graphics in our Forum, you have to make liberal and judicious use of the English language. It's the only vehicle we have to understand what you propose to do. For example:

1) You state you don't want to use a fired heater upstream; fine, but what is it that you are going to use to heat up the feed to the asphalt unit? Presumedly you are going to use a preheat exchanger train (which you say you will make more "efficient").

2) Then you ask about flashing the feed in the exchangers (the same preheat ones?) inspite of the inferior heat transfer coefficient you know you are going to get. This doesn't make chemical engineering sense; how can the exchangers be more efficient with a rotten film coeffient?

3) If you have a backpressure control valve downstream of the preheaters you don't have to flash in the exchangers. That is, unless the valve is not properly sized or not working correctly. But you don't discuss this nor state why it should only work "to a certain extent". This also doesn't make process sense.

The biggest reason not to flash in the exchangers (besides ruining any heat transfer efficiency and risking metallic erosion) is that you are now in the "great & wonderful world" of 2-phase fluid flow where things can only get bad instead of staying stable and with low pressure drops. Your pressure drop will increase and since you admit that your feed pressure is limited, then your efficiency on flashing is only going to get worse. I would do anything to avoid 2-phase flow in equipment whereever I could; otherwise, you're asking for trouble and more problems.

Perhaps if you give us all the scope of what you are doing with specifics, we can make better sense out of this proposal; but with what you've given, it doesn't seem logical. Sorry, but I'm trying to understand what you've written with 44 years of experience and I'm having trouble. I hope I've mananged to make some sense myself.

Art Montemayor
Spring, TX

 

[Rosalynn]

The existing preheat train first heats the crude (enters at 80C or so) with a side draw (VGO at about 200C) from the asphalt column. Then the feed is further heated with the asphalt from the bottom of the asphalt stripper (300+C). The LMTD's in these exchangers are in the 70-80C range. The feed to the flash vessel gets up to only 160 or 170C, and I would like it to be in the 200-210C range. I believe some more exchanger area would be economically justifiable, say LMTD's of 40-50C. The duties are available in the rundown streams--both exit the unit hotter than required. This is what I mean by a more efficient heat exchanger train. If I can take the duty out of a hotter-than-desired rundown stream and use it to heat up my feed stream, that seems to me to be a more efficient use of energy--even if I might need a bit more area to get heat transfer because I have some vapour present.

I've modeled the unit in Hysys. There is little flashing in the exchangers at 160 or 170, but it is significant when the temperature rises above 200C (yes, in the same exchanger train--but with the addition of one or two larger exchangers).

My upstream supply pressure is limited by the pressure in the wells producing upstream and by the required separation pressure in upstream vessels. Yes, I could recommend pumps to boost the pressure, but would I do that just to prevent flashing in exchangers? It also raises the concern of exceeding the design pressures of the piping and existing exchangers. I couldn't increase by much before reaching those limits in a blocked outlet situation.

I mentioned that there is a backpressure controller because it would be an obvious solution if one wasn't present. We have one and it works fine. My point was that increasing its setpoint enough to prevent flashing at 200+C feed temperature would likely cause it to pinch in so much that we couldn't get the throughput we need.

Some additional information:
Feed pressure to unit: about 550 kPag
Feed temperature: about 80C
Pressure controller upstream pressure: about 180 kPag
Flash vessel pressure: 30 kPag

Montemayor's point about risking metallic erosion is at the heart of my concern. There are actually three pitch/feed exchangers (identical exchangers in series). The feed is on the shell side (AET type; I have the spec sheets if construction info is needed). These are currently the last exchangers before the feed enters the flash drum, and I am concerned about causing damage or vapourlocking the last one in line if I add more exchanger area upstream.

I appreciate the quick response to my post and the offer of help. Not so much the implication that my thought processes are illogical, though. Hopefully this additional info helps to clear things up without adding more confusion.

 

[Montemayor]

Rosalynn:

You must be trying to read between the lines when you state that I have made an implication that your thought processes are illogical. Your thought processes are just fine - as evidenced when you furnish a clear and detailed scope or work. I made no comment of your thought processes; but rest assured I certainly would if I had found them lacking as I did your first process description. We are engineers and we can only get to the bottom of problems with facts and reality.

Your concern, in my opinion about the backpressure valve is not merited unless you have a very constrained Cv value for the valve. If you designed it correctly, "pinching" it too much should not be a concern. The Cv value is what determines the throughput.

I still maintain you're overlooking a major bottleneck if you do not simulate for the 2-phase flow you will get in the exchangers. You have not stated you've accounted for it, so I have to assume that you haven't considered the effects it can have in your pressure drop.

I hope these observations are of some help.
Good Luck. Your logic in seeking other heat sources for the preheat is the proper, logical path to follow as a first preference. You're on the right thinking path.



Art Montemayor
Spring, TX

 

[jalvarez]

Rosalynn, a couple of hints:
1.- At 220ºC you have the possibility of increasing the pressure for class 150 piping up to 1200 kPa. It seems that you have enough room to do something with your pumps. Check if you may use bigger impellers, or add a booster pump, low DP. (I don't like booster pumps, but sometimes they solve the problems).
2.- in designing a possible additional heat exhanger/s, be very careful with the addition of seal strips. Sometimes they are missed in the process design, or in the jump to mechanical design. And this is a big mistake, as they are the solution for viscous products like these.
Have a safe day.
J. Alvarez

 

[Rosalynn]

Thanks to both Montemayor and J. Alvarez. Input much appreciated, and will keep all points in mind as I move forward.
Regards,
R

 

[25362]

The last heat exchanger in a train frequently flashes the liquid into a tower or drum. The temperatures in the HE reach a peak and then drop due to flashing (latent heat) improving somewhat the LMTD. The OHTC should improve (not deteriorate) by some 30% in the vaporizing section of the exchanger. Thus, it is possible that the use of a back-pressure valve wouldn't help but, on the contrary, reduce the heat transfer capabilities of the last exchanger.

Since the friction drop on two-phase flow should be higher -as mentioned by Montemayor- the number of tubes per pass and/or their diameter is increased at the vaporizing end to cope with this phenomenon. This kind of design is very common in plants vaporizing solvents, such as, for example, lube extraction and solvent dewaxing units.