12 replies to this topic

[karthik.kattupalyam]

Dear members,

 

i am working as a trainee engineer,

 

I want to increase the flow rate of feeding into the reactor from 1800 kg/hr to min 2800 kg/hr(hydro toluene) because to increase the production. so i use graphite heat ex-changer (80 m2) for circulation to maintain the temperature below 28 deg C to process requirement. my problem is if i increase the flow rate from 1800 to 2300 kg/hr i cant maintain that temp below 28 deg c, actual i am using 50 m3 pump for circulation and brine chilling for graphite heat ex-changer its temp around 8-9 deg c.

 

 

Data:

vessel GLR 12kl, jacketed for cooling (brine chilling is using 8-9 deg C )

50m3/hr pump for circulation(s.s)

feeding:

1.hydro toluene(density 0.945 gm/cc feeding temp around 22-30 deg c from heat ex-changer)

Hydro toluene is just name (actual its coming from hydrogenation of ONCB, & it have toluene also)

2. H2SO4 spent 62%(feeding temp around 35-40 deg c from heat ex-changer). feeding rate should be 73% of Hydro toluene.

 

 

 

my calculation is 

 

I calculated like Q=U A Delt T(lmtd)= 500*80*(15)=600000 kcal/hr, 

50m3/hr means 50,000 lit/hr it is pumping very low flow. 

so i think this heat ex-changer is much enough, maybe  i have to increase the pump cap around 100 or 150 m3/hr

 

so please tell is it right or any other solution is there, it will help to increase  my production

Edited by karthik.kattupalyam, 20 September 2013 - 06:29 AM.

[Dazzler]

Hi Karthik,

It seems the heat exchanger is there to remove heat generated by a reaction or mixing.  Perhaps you can calculate the heat removal duty for the current operation and then work it out for the new target operation.  Then model more rigorously the heat exchanger to "rate it" and see what it is capable of and what parameters you can feasilby change to improve the heat removal.

 

Keep in mind that heat exchangers can foul so if it is currently not perforning as well as original design (find a process datasheet) then it might need cleaning.

 

Also, the flow through the exchanger may have been set to achieve recommended velocities for a graphite heat exchanger.  Perhaps you can research whether there is a maximum velocity for such materials of construction so the graphite is not erroded away.  In addition, if the exchanger has tubes, increasing the flow/velocity might increase the change of vibration that could lead to breakage. 

 

Dazzler

Edited by Dazzler, 21 September 2013 - 07:58 PM.

[curious_cat]

What's the heat load you need? Is this a heat of reaction? A heat of mixing? How many kcal/hr. 

 

You did a HEX rating calculation. Fine. But first, do a process side heat duty estimation. 

 

"so i think this heat ex-changer is much enough," is meaningless without a benchmark to compare to. 

[fseipel]

Is reactor fitted with condenser?  The most effective form of cooling in my experience, is to reflux the solvent to remove the heat of reaction.  Depending upon its boiling point and whether you have vacuum available, this may or may not be possible; if ONCB is what I think it is, it is a very high boiler; however any toluene present would reflux at modest vacuum and provide exceptional cooling, much more effective than liquid-liquid heat transfer since you would have condensation on one side of H/X.  If not possible, I noticed you have a stainless pump, if that is holding up/no corrosion issues, why not use a stainless heat exchanger which is cheaper than graphite?  How is SS pump even holding up in this kind of sulfuric acid mixture?

 

Make sure GLR jacket is fitted with agitated nozzles on the jacket and jacket flow exceeds minimums.  Also inspect jacket for corrosion.  However, jacket is probably not providing that much of your overall cooling (very poor heat transfer coefficient due to thick steel + corrosion + insulating layer of glass + low velocity).  Consider a Hastelloy heat exchanger or a plate & frame graphite exchanger instead of shell & tube.  If you have a GLR with split pipe jacket it will work better than solid wall jacket for cooling.

 

Since it's already running, obtain empirical data rather than theoretical heat of reaction. Shut off the cooling for a minute or two, and feeds, and monitor empirical temperature rise (Delta T).  Then calculate heat using Q=m*CP*Delta T using CP = composition weighted heat capacity of reactor, m=mass of charge.  This will tell you how much heat you must remove to maintain temperature.  Do this at beginning or whenever rate is maximum.

 

Next measure flow rate, inlet outlet temperatures on process side or brine side of H/X and solve Q = m*CP*Delta T using m = mass flow rate through HX and Delta T = temp. change across exchanger.  Then solve for U using that Q.  Consult with manufacturer if this 'U' is reasonable.

Consider pre-chilling reactants.  Presumably you have 'stir out' time at end of reaction and 'transfer time' to pump down reactor and other downtime (weekends, etc).  During such times, use the storage tank (H2SO4, OCNB) transfer pumps, to instead recirculate those tanks through a small heat exchanger, back to storage tanks, or insert cooling coils in feed tanks.  This will also provide a more uniform load on your chiller so you don't 'shock' it with surge loads.  Alternately with creative piping it may be possible to use the recirculation exchanger for such prechilling (i.e. whenever reaction isn't running or when it is being pumped down/out).

 

As for pumping rates, report back what the approach temperature is between brine inlet temp & process outlet temperature.  That will give some clue whether you need more area, or a higher pumping rate.  It may also make sense to look at the number of exchanger passes and pressure drop.

Edited by fseipel, 22 September 2013 - 03:38 PM.

[curious_cat]

How is SS pump even holding up in this kind of sulfuric acid mixture?

 

 

My guess: It's teflon lined SS for the pump

 

Make sure GLR jacket is fitted with agitated nozzles 

. 

Alternatively, a spiral jacket insert / baffle. 

 

 

 

 Alternately with creative piping it may be possible to use the recirculation exchanger for such prechilling 

 

It's a good idea but often a operational hazard depending on the consequences of the wrong valves being manually opened. 

Edited by curious_cat, 22 September 2013 - 03:46 PM.

[fseipel]

 

Alternatively, a spiral jacket insert / baffle.

 

I'm not familiar with this.  Do you mean a half pipe or split pipe jacket as suggested in my response or something different?

 

 

Alternately with creative piping it may be possible to use the recirculation exchanger for such prechilling 

 

It's a good idea but often a operational hazard depending on the consequences of the wrong valves being manually opened.

Yes, pumping reactants back to the feed tank could pose a hazard or at least a contamination issue.

 

The recirc exchanger itself can also pose hazards -- I've seen concentrated acid leak back to boilers/chillers from split heating/cooling, and also, glycol solution leak from tube sheet into water reactive materials.  In this case it's probably a graphite block style exchanger.  Measuring conductivity of the brine with a transmitter/alarm may merit consideration also.

 

If a condenser and modest vacuum system are available and fitted, reflux cooling may be the least expensive option since no new equipment is required.  This effectively uses the toluene (or other solvent) as a refrigerant, without a compressor.  As toluene is condensed, it will drop back into pot, and cycle repeat.  The pot temperature is controlled by regulating the operating pressure/vacuum and/or coolant flow/temperature to condenser.  It is even more effective if you have direct expansion on the coolant side.

[curious_cat]

 

I'm not familiar with this.  Do you mean a half pipe or split pipe jacket as suggested in my response or something different?

Different. It is a thin strip welded either to inside of jacket or outside of vessel in a spiral manner. It channels flow down the jacket somewhat like a pipe. In theory at least. 

 

 

 

It is even more effective if you have direct expansion on the coolant side.

What do you mean by this?

[karthik.kattupalyam]

 

How is SS pump even holding up in this kind of sulfuric acid mixture?

 

 

My guess: It's teflon lined SS for the pump

 

Yeah it is teflon lined ss.

[karthik.kattupalyam]

What's the heat load you need? Is this a heat of reaction? A heat of mixing? How many kcal/hr. 

 

heat of mixing

 

[curious_cat]

Why don't you estimate how much is this heat of mixing load then? 

 

What makes you think you can rate / debottleneck your HEX without knowing that?

[Padmakar Katre]

Dear members,

 

i am working as a trainee engineer,

 

I want to increase the flow rate of feeding into the reactor from 1800 kg/hr to min 2800 kg/hr(hydro toluene) because to increase the production. so i use graphite heat ex-changer (80 m2) for circulation to maintain the temperature below 28 deg C to process requirement. my problem is if i increase the flow rate from 1800 to 2300 kg/hr i cant maintain that temp below 28 deg c, actual i am using 50 m3 pump for circulation and brine chilling for graphite heat ex-changer its temp around 8-9 deg c.

 

 

Data:

vessel GLR 12kl, jacketed for cooling (brine chilling is using 8-9 deg C )

50m3/hr pump for circulation(s.s)

feeding:

1.hydro toluene(density 0.945 gm/cc feeding temp around 22-30 deg c from heat ex-changer)

Hydro toluene is just name (actual its coming from hydrogenation of ONCB, & it have toluene also)

2. H2SO4 spent 62%(feeding temp around 35-40 deg c from heat ex-changer). feeding rate should be 73% of Hydro toluene.

 

 

 

my calculation is 

 

I calculated like Q=U A Delt T(lmtd)= 500*80*(15)=600000 kcal/hr, 

50m3/hr means 50,000 lit/hr it is pumping very low flow. 

so i think this heat ex-changer is much enough, maybe  i have to increase the pump cap around 100 or 150 m3/hr

 

so please tell is it right or any other solution is there, it will help to increase  my production

Hello,

From your post the exact cooling system configuration is not clear. Please upload a simple sketch with the P,T and Flow conditions.

For duty calculations you have assumed overall HTC "U" as 500 kcal/h-m2-C, how you have calculated. There are certain guidelines for estimation of jacketed vessel HTCs. It needs lot of process data and geometry data as well.

Please reply point wise to my questions,

1. What is agitator type if present and it's details such as speed, diameter, number of blades (if PBT)

2. What are typical clearance for agitator

3. Are you accounting diccipated heat due to agitation while calculating the overall heat load on cooling system

4. Is the heat transfer surface arrangement - a convention jacket if so is it baffled/unbaffled

5. What type jacketed fluid entry, radial/tangetial

6. Looking at the increase in throughput and thereby increased brine rate of ~150 m3/h from 50 m3/h (have you checked the pump adequacy?) will lead to extra pressure drop in jacket and how you are planning to solve this issue?

 

Please note that existing hardware has fixed heat transfer area, increase in flow rate will not lead to significant increase in HTC. So you have to evaluate the case carefully.

 

P.S. - Search on internet with keywords such as, Jacketed vessel heat transfer and you will get lot of good stuff to understand the basics involved before you actually start doing calculations.

[fseipel]

 

Quote

 

 

It is even more effective if you have direct expansion on the coolant side.

What do you mean by this?

 

Direct expansion refrigeration -- wherein you run the liquid refrigerant line out to the condenser.  So the HCFC refrigerant vaporizes in the condenser.  Essentially, instead of having an evaporator chill brine water or a water/glycol solution, the refrigerant evaporator, IS the process condenser.  The advantage of this, is spectacular heat transfer; the disadvantage, is that separate compressors are required for separate reactor systems and oil/refrigerant separation can pose a problem if not designed properly.  Also a long suction line may be required from condenser back to compressor unless it can be mounted close to condenser.  The advantage is improved heat transfer.

 

 

I'm not familiar with this.  Do you mean a half pipe or split pipe jacket as suggested in my response or something different?

Different. It is a thin strip welded either to inside of jacket or outside of vessel in a spiral manner. It channels flow down the jacket somewhat like a pipe. In theory at least.

I'm interested in learning more; I've heard of Optimix from DeDietrich but this sounds different.  Who supplies it?

Edited by fseipel, 23 September 2013 - 03:40 PM.

[curious_cat]

Direct refrigeration sounds very interesting! I've never seen it before. Thanks.

 

For spiral baffles see here:

 

http://books.google.... baffle&f=false

 

http://books.google.... baffle&f=false

 

I'm not sure about which of the standard vendors sells this. Ours were smaller custom fabricated or small-vendor items.