Tuesday, 23 January 2018

Guidelines for Process Development, Optimization & Scale up

Dear All.....!!! Good morning everyone....!!!

Its been a long time since i've posted here, and it is because of my busy schedules,

Since the startup of this site, i'm welcoming new guys to contribute on this site but till now i haven't received any requests except one, and since the one is not having enough knowledge about blogging i'm posting behalf on that guy, and his name is Mr. Saravana Kumar B, he is colleague since one year. Being a process engineer, he has sound knowledge on Solvent Recovery and Process Development. 




Lets start the show: 


Posts you may like:
[How to] reduce vapour losses during distillation ? 
[How to] Select filtration equipment ?


Role of cTSD: 

1.  To help R&D develop processes that are safe, environmentally manageable, and that can be scaled up and implanted to get reproducible production with consistent quality and efficiency.  (Phase 1)
2.  To establish proper scale up of  the process. (Phase 2)
3. To help production department  take over the process  for successful commercialization. (Phase 3)




To fulfill these roles c TSD process engineers will use various guidelines. Knowing these  will remove misunderstanding and also avoid rework, loss of time and will help in identifying use and availability of proper equipments and materials (raw materials) . These guidelines will be useful for R&D work as well as during scale up. These guidelines are not exhaustive and these will be revised , improved from time to time.

These guidelines are classified as follows:

1.     Process
2.     Equipments


Each one is discussed below:

A.    Process: 

1.     Solvents:

1.     If there is option we should avoid use of very low boiling solvents (BP<60 C). Low boiling solvents will require use of brine (high energy and capital cost). Losses are generally high in low boiling solvents and also cause environmental pollution.

2.     We must try to recover and reuse the solvents in the process. It may not be possible to dispose off used solvents in future (environment enforcement) and burning them will be also very expensive. Cost of incineration is high. Also chlorinated solvents cannot be burned (easily) and lead to additional pollution (Dioxins, PCBs)

3.     Also use as less solvent as possible as higher usage means higher losses and bigger equipment. If the solubility of compound at normal boiling point is limiting try at higher pressure (which will give temperature above normal BP) rather than increasing the solvent quantity.




4.     Avoid use of mixed solvents or change of solvents from step to step.

5.     If choice exists use water immiscible solvents as this avoids pollution (COD) due to soluble solvents.

b. Reaction Conditions: 

1.     Temperature: Normal range to choose without requirement of special utilities like hot oil / high pressure steam or brine is between 40 to 135 C. Do not say room temperature as a condition as room temperature can vary between 15 to 42 C.

If the product is heat sensitive (better to assume all API as heat sensitive) what will matter is not only reaction temperature but also jacket heating media temperature. Therefore selection of jacket temperature level will be important and it is best to specify D T. Have a method to measure jacket temperatures in lab and plant. Any D T higher than 10 C should be evaluated with caution.




The controllability of temperature in plant environment will depend on number of factors. It is therefore necessary to specify the range of control e.g. 60 ± 2. This will decide the heating / cooling media temperature and flow rate and control strategy.

Rates of reaction double with every 10 C increase in temperature. If some reactions are slow (taking 10 – 40 hours) try reaction at higher temperature if permissible. To get higher than atmospheric boiling point temperature consider operation under pressure. (You may need lab autoclave for this)

It is important to understand TTT concept. This stands for Time, Temperature and Transformation. Transformation is either desirable or undesirable chemical reaction. When we subject chemicals to high temperature for short time this is equivalent to subjecting them at low temperature for more time. Many times we will actually get less impurities if we carry out reaction at higher temperature in short time. (Depending on relative kinetics of main and side reactions.)


2.     Homogenous v/s heterogeneous reactions: 

Reaction where from beginning to end the reaction mass in one liquid phase and also where during course of reaction no gas or solid is added or formed is homogeneous reaction. These are the most preferred from scale up point of view.

Heterogeneous reactions involving two or more phases (liquid / liquid, Liquid /Gas, Liquid / solid or all three together); either added or formed during reaction will spring surprises in scale up. The type of agitator and scale of agitation will be important. Chemical Engineers will guide in this regard. Do not use general purpose ANCHOR for all types of reactions. The particles size and distribution in case of solids formation will depend on type and speed of agitator.




For reactions involving gas as raw material the design of sparger will be also important in addition to agitator for successful scale up. Single point dip pipes are generally not adequate.


3.     Reactor filling:

Reactor (whether in lab or pilot plat or plant) if under filled or overfilled at any stage in the reaction cycle is not acceptable. Typically filling should be between 60 to 85 %. This will decide the batch size for a given reactor or decide a reactor volume for given batch size.

Both under filling and overfilling do not give proper agitation. Under filled reactor is also bad as reaction material is exposed to hot jacket surface inside the reactor and this can degrade / decompose. 




If during the course of operation the volume is going to fluctuate (as in the case of distilling out one solvent before second solvent is added where change of solvent is necessary), use of other equipments like falling / agitated thin film evaporators must be used (even in the lab scale). We can also use reactor where we continuously make up feed as the level drops, this way the filling of distillation reactor is maintained at desired value.

Use of split jacket will be also necessary to handle varying volumes in the reactors.

4.     Reactions under pressure:  It is quite safe to carry out reactions under pressure. This allows using higher temperature.

5.     Reactions under vacuum: Always measure vacuum on the reactor by Mercury manometer (and not by Gauge on the vacuum pump). Always perform leak test before starting every batch. Have mechanical seals on agitator, and break vacuum with N2 when solvents are used.

6.     Jacket heating / cooling media and condenser   cooling media:

These heating cooling fluids (water/ steam/ hot oil / brine /kerosene etc) 
Must be compatible with the process over entire range of reaction and at reaction conditions. (Hazop will bring these issues). Use of fragile glass condensers should be avoided when the coolant has poor compatibility with the reaction mass.

7.     Material of construction of equipments in contact with process:

This will be selected by process engineers based on experience, literature and actual corrosion study. Corrosion of equipments can alter the color and purity of products in addition to the danger of equipment failure.

8.     Run away and exothermic reactions: In such cases all reactants should never be added and the limiting reactant must be added slowly only after ensuring that there is no accumulation of un reacted material in the process. Suitable analytical methods have to be developed for this.

9.     Handling of solids: It is far more easier to handle liquids and solutions than wet / dry solids in the plant. Attempts should be made to dissolve solids (both RM and Intermediates) in solution and the added to the reactions. We can even dissolve wet cake in the solvent used in the next step and dry / dehydrate in the solvent instead of as dry powder. It is also possible to dissolve wet cake in the filter itself (ANF) and transferred as solution (or slurry.). It is possible to dissolve the cake on filters and centrifuge using the solvent used in the next step.

10.   Reflux during reaction:  There is lot of misconception about reflux. If during no water and gas is to be removed, then the extent of reflux is not important at all. In fact reaction rate in all such cases depends only on temperature and even just boiling temperature with zero reflux will give same results. Unnecessary reflux means waste of heating / cooling energy as well as loss of solvent through condenser.

11.   Reaction Maintenance time: It is always better to follow reaction with suitable analysis to ensure that it is terminated as soon as it is over. In fact for all reactions a control test is a must




12.   Mole ratios: Large molar excesses should be avoided as these result in to generation of waste and increase cost of products and waste disposal. Any reactant with use of more than 10 % molar excess over theory should be critically evaluated.

13.   Purity of raw materials and solvents and intermediates:  These must be established well in.  Advance and should reflect commercially available specs and not high purity lab chemicals specs. Use of recovered products and solvents should be included along with the appropriate specs

14.   Stirrability and pumpability of reaction mass:  It is necessary that during entire reaction period the reaction mass remains stirrable and pumpable. The temperature and solvent quantities should be decided on this. This problem can be frustrating during scale up.

c.      Reaction workup:


1.     Distillation / removal of solvent:  Pl refers to point 3 (reactor filling). In addition to above point of filling it is important to know when the solids are coming out, how this is affecting stirrability, the elevation of boiling point as distillation proceeds, D T between jacket and process, Measurement and control of vacuum is necessary in vacuum distillation operations. For solvent removal falling film or ATFE are best as they give very small (minutes) of exposure to high temperature as compared to hours in reactor distillation process. Any product degradation will be reduced substantially.

Never remove all solvent leaving very small quantity in the reactor, which will be difficult to stir. Such process will be difficult to scale up.

2.     Cooling: Here also note the cooling rate, D T, point at which solids (if any) coming out, stirrability / pumpability. If solids are coming out then agitator has to be turbine and not anchor.

3.     Filtrations: The slurry density and flowability are important. Typically do not plan slurry density (qtty of free solids per kgs of total mass) beyond 25 %.

If the filtration temperature is below 20 and above 40 C, then filtration equipment will have to jacketed and insulated and suitably heated or cooled. Even in lab this will have to be done, otherwise the slurry temperature before filtration will be ok, but will change substantially during filtration giving wrong data / result.

It is necessary to build sufficient cake thickness (40 mm min) on lab Buchner funnel to understand the filtration characteristics. (Pl notes that filtration time is proportional to square of cake thickness. In plant if cake thickness is say 5 times of lab then the filtration time will be 25 times!)

Washing is important to get purity by driving all ML, which contains impurities. The practice of re-slurring in centrifuge is bad and cannot be reproducible. The quantity and temperature of wash solvent is important. It is necessary to develop a control test to set impurity limit in the last wash to ensure effective washing.


4.     Extractions / washing and layer separation:

Typically here some washing with water or organic solvent is done to dissolve / extract soluble component. This is followed by layer separation. For good extraction, which is heterogeneous process, proper agitator (Turbine) with baffled reactor is required and anchor is poor choice. Also location of turbine in liquid is important.  The quantity of first wash should be 5 % more than required making saturated solution; this helps in reducing (effluent in case of water) volumes. It is better to give more no of small volume washes than less no of washes of large volume.

Ease of layer separation depends on density difference between aqueous and organic layers. The deliberate higher differences can be created by manipulating solvent/water volumes or even by adding extra salts (say NaCl). Other important issue in layer separation is emulsion formation. This is due to some surface-active impurities and suspended matter. Emulsion can be broken by electrolytes, higher temperature, filtration, centrifugation etc. Sharp separation will improve yield, reduce effluent COD, and improve purity of both layers.


5.     Neutralization / basification with alkali:  If final pH is important then it is better to use as dilute alkali as possible (say 5% NaOH), we can also use milder alkali solution like Na2CO3, NaHCO3 etc. Also this is likely to be two phase (heterogeneous process) so agitator considerations will apply.

6.     Neutralization / acidification with acid:  If final pH is important then it is better to use as dilute acid as possible (say 5% H2SO4), we can also use milder organic acids if permissible .Almost all dilute inorganic acids (except HNO3) attack stainless steel and chloride also attacks SS. The mode of addition and choice of acid, concentration and final pH are very important. Also this is likely to be two phase (heterogeneous process) so agitator considerations will apply.
B.    Equipment:

Equipment will have to be selected from operatibility, safety and environment considerations. Broad classification is done below for guidance for few common  equipment required  for post reaction operations.

1.     Filtration equipment:
Preference
Equipment
Cost
SHE angle
Scale up angle
4
Centrifuge top discharge Standard.
Low
Bad, exposure to solvents and material, fire hazard.
Manual Washing can be unreliable.
2
Centrifuge bag lifting type
Low
Fair. Less exposure to solvents and material. Fire risk is there
Manual Washing can be unreliable.
1
Centrifuge bottom discharge with N2 pressure
High
V Good, No fire and exposure hazard
Entire cycle from filtration to washing is automated. Reliable scale up.
1
Reversing bag centrifuge
V. High
Safest. No residual cake after discharge. Ideal for low volume high price products.
Entire cycle from filtration to washing is automated. Reliable scale up
1
Plate and Frame filters
Moderate
Minimum exposure to solvents and product discharge, Fire risk less as air is not sucked. Less solvent loss to environment.
Very good for scale up as cake thickness is low. Good washing can be done. Can handle large qtty and reactor is freed in one go.
2
ANF
High
Minimum exposure to solvents and product discharge, Fire risk less as air is not sucked. Less solvent loss to environment.
Very good for scale up. Good washing can be done. Slurry can be transferred and solids handling can be avoided.
3
Closed pressure nutche
Low
Exposure while discharging the product.
Not reliable as cake may crack and washing will be ineffective.
4
Open vacuum Nutche
V low
Very bad as exposure to solvents and solid
Washing can be problematic depending on size and nature of product.

Larox
V high
No exposure, completely automated operation. May be good for aqueous systems
Reliable scale up. Possible to get low residual moisture.



























































2. Drying equipments:

Preference
Equipment
Cost
SHE angle
Scale up angle





3
Tray dryer
Low
Poor. Exposure to product and solvent going to environment.
Fire hazard due to dust
Poor. The temperature and airflow inside dryer is complex. Damper position is also important.
2
FBD
Moderate
Poor. Exposure to product and solvent going to environment.
Fire hazard due to dust
Moderate.          
1
Double Cone Vacuum
Moderate
V Good. Allows solvent recovery. Exposure to product is less.                       
Good, If vacuum and temp are maintained.
2
Vacuum Tray dryer
Moderate
Fair. Solvent can be recovered. Exposure to product is there.
Moderate.

Continuous flash dryer
Moderate
Fair, Solvent lost to environment. Exposure to product low.
V Good. Avoids manual handling. Ideal for large volume production. (Ibu / Cipro)
































3.    Layer separation


Preference
Equipment
Cost
SHE angle
Scale up angle





3
Gravity separator manual
Low
Poor separation, emulsion layer, manual judgment
Moderate, Filtration prior to layer separation helps in breaking emulsion.
2
Gravity separator automatic
Low
No error due to manual judgment. Much superior to manual.
Moderate, Filtration prior to layer separation helps in breaking emulsion.
1
Centrifugal separator
High
Excellent. Sharp separations. No exposure
Good

4.     Concentration / solvent removal


Preference
Equipment
Cost
SHE angle
Scale up angle
3
Reactor
Moderate
No issue if condenser is adequate
Not good. Material degradation
1
ATFE
High
No issue if condenser is adequate
Excellent. Exposure of material at distillation temp only minutes.
2
FFE
Low
No issue if condenser is adequate
Excellent if solids do not come out. Minimum product damage.


That's it......!!! These are the most basic and reliable techniques used during process development, optimization and scaleup.

Hope the above mentioned are understandable and are upto the mark.

Hope he will be back with a new post......!!!!

Feel free to comment, comments are most appreciated.....!!!


Related Posts:
[How to] Perform Material Balance calculation  

What is LMTD correction factor ?
[How to] Calculate required vent sizing of vessels ?  
Selection of Equipment for a manufacturing process 


About The Author


Hi! I am Ajay Kumar Kalva, Currently serving as the CEO of this site, a tech geek by passion, and a chemical process engineer by profession, i'm interested in writing articles regarding technology, hacking and pharma technology.
Follow Me on Twitter AjaySpectator & Computer Innovations


10 comments:

  1. Dear friend, good morning....im designing column reactor , having 40 inch height & 30 inch diameter. having two input nozzle at the bottom side . one nozzle having sparger inside reactor which having flow 2-5 lt/ min(MDC 40°C) & other nozzle also bottom but which is tangential to the bottom having hot water 85-90°C. this process is carried in Hot condition . please tell me what capacity of heat exchanger required? vapor column height & diameter size?

    ReplyDelete
  2. at height 34 inch there is 2 inch overflow nozzle for product, this line goes to centrifuge.

    ReplyDelete
  3. this is continuous process...

    ReplyDelete
  4. Dear Ajay,
    In workup stage we are giving no of wash ...is it any calculation for the same?? How it can be calculated??

    ReplyDelete
    Replies
    1. Dear Mahesh,

      Surely we will have a calculation for the washing solvent qty., it will be calculated based on the formed product and formed bi-products, based on their solubility in the washing solvent, the washing qty. shall be derived.

      Lets suppose, we are having a reaction where Soda ash is produced as bi product, we need to remove it by water washing.
      Soda ash solubility is 0.5 Kg/L(Say),
      And in the reaction the subject bi-product formed will be 50 Kgs,
      then the amount of water required is 100 L,
      So washing qty will be 100 L, and based on lab trend if the subject is getting washed out 50% in each washing, then approximately 3 washing required each 70 L, 50 L, 40 L,

      I've considered excess qty. in each washing as safety purpose(if the product is insoluble in water).

      Finally we need to end up with a brine washing(to remove the water traces from the product layer)

      Hope you understood the trick.

      Regards,
      AJAY K

      Delete
  5. Dear Ajay,
    If we want to maintain 1.5kg/cm2 jacket pressure with utilities like cooling water or brine how much will be flow requirement or line size??how it can be calculate??

    ReplyDelete
    Replies
    1. Dear Mahesh,

      Pressure and flowrate aren't having any direct relationship, so it is not possible to explain it,

      Let me give you one simple example of what i've mentioned above,
      Let's suppose 100 L/hr flow is there in a line, the line is throttled at some point and after the point if you can measure the pressure it will be high when compared to previous line, but the flow may remain same / may get reduced.

      If pressure is increasing means then the flow is not upto the mark, Pressure exerts in all directions, what we measure is the pressure exerts on the sides of the line, and we expect is the pressure that is exerted on the forward element :) .

      Hope you understand

      Regards,
      AJAY K

      Delete
  6. Dear Ajay,
    If I want to scale up batch from pilot to commercial scale ,if pilot reactor having RPM 55 with 4 baffles but commerical reactor having no baffles.so how to considered baffel effect during RPM calculations for Reaction, Crystallisation & Workup stages ????

    ReplyDelete
    Replies
    1. Dear Mahesh,

      Calculate the reynolds no., then consider 10% of excess turbulence with 4 baffles and 4% excess with 2 baffles and 7% excess with 3 baffles. This is a thumb rule.

      For reaction the RPM should be vigorous and for workups it depends[it depends up[on the formation of emulsion, if emulsion is formed with higher RPM, then reduce the RPM based on pilot plant experience], for crystallization go with mild RPM to avoid crystal breaking.

      Regards,
      AJAY K

      Delete