How Process Optimization Cut High Chemical Production Costs

Due to the interest on the recent post on how a major dye and chemical company improved process safety and shortened downtime through process modification, this post will discuss how a major specialty chemicals manufacturer used process optimization to:

  • Cut solvent consumption by 50%
  • Decrease production costs by 10%
  • Reduce batch time from 10 hours to 4 hours


The Challenge

METTLER TOLEDO RC1 Reaction Calorimeter

Figure 1: Substitution reaction according to procedure in current use (RC1e experiment)

A major speciality chemicals manufacturer sought to reduce the high costs of solvent consumption in their current production process.  The problem involves both high recycling costs and evaporative losses to the environment. The potentially hazardous nature of the process made in-plant optimization difficult. The plant in present use includes extensive safety measures to prevent a thermal runaway in case of equipment failure.  A Reaction Calorimeter was used in attempt to reduce solvent costs while maintaining process safety as well as product quality and yield.

The Procedure
The existing procedure was first simulated in the RC1e Reaction Calorimeter (Figure 1).  While heating the reactor content of raw material A to 40 °C, a stoichiometric portion of reaction partner B is added. The reaction mass is held isothermally until the heat production has decreased.

process safety

Figure 2: DSC curve of the final reaction mass

In order to accelerate the reaction, an excess of reactant B is then added and the reaction mass is again kept at 40 °C. Finally, the temperature is raised to 46 °C for the reaction to run to completion.  The heat production curve showed that only about 10% of the feed was converted at the end of the first addition. If a malfunction – such as cooling failure or stirrer breakdown – were to occur at this stage, the continued reaction of the accumulated amount of B would cause a runaway without the use of special emergency measures.

The reaction mass would heat up leading to possible secondary reactions that can be estimated from the microcalorimetric study (DSC curve, Figure 2). In the temperature range 100 to 200 °C, an exothermic reaction of -290 kJ/kg takes place. At 300 °C, a violent decomposition then sets in with a heat production of another -220 kJ/kg. This is the particular reason why the safety of this process was considered critical. Any changes in the process must thus first be checked thoroughly by measurements to prevent a trade of lower production costs for an increased process safety risk.

Modification of the Procedure
The analysis of the existing procedure suggests a semi-batch process with dispensing rate, reaction temperature and concentration of reactants was investigated.  Due to the reproducible operation of the RC1e reaction calorimeter, a wealth of data was acquired from each individual experiment.

METTLER TOLEDO RC1e Reaction Calorimeter

New procedure run at 62 °C isothermally and at higher concentrations (RC1e experiment)

The data was then used to design anoptimum procedure (Figure 3). The reaction is now performed at 62 °C (boiling point of the solvent). The higher temperature and the selected addition time resultedin a considerably enhanced reaction rate. The increased heat production of the reaction can be dissipated without problem through efficient use of evaporation cooling. In addition, the faster reaction keeps the accumulation of nonreacted material now below 0.4 = 40%. The resulting adiabatic temperature rise was found to be at a non-critical level.

Realized Benefits
The solvent consumption is cut by half. Total production cost savings amount to 10%. Since the reaction runs considerably faster, the cycle time
has been reduced from 10 hours to 4 hours.  Although the primary goal did not deal with improving process safety, an appreciably safer design of the process was achieved owing to the exact knowledge of the process data.  As a consequence, certain costly emergency measures were rendered unnecessary.