Abstract
The reaction and separation systems of a continuous chemical process can be considered independently to identify alternatives.  Interactions between the two systems are then taken into account to determine potentially feasible process flow diagrams.  A subset or cluster of potentially inherently cleaner process options can then be selected for further consideration using economic, safety and environmental criteria.  In this paper, a prototype screening methodology is presented to facilitate the independent identification of the potentially feasible alternatives for the separation system.  The current capabilities of the presented prototype, its relationship with existing approaches and key limitations are described.

Introduction
Reaction and separation systems are typically designed sequentially (Mizsey & Fonyo 1990, Smith & Linnhoff 1988).  To minimize process development time, process conditions in the reaction system are often "frozen" to optimize selectivity and conversion (Ingleby et al. 1986).  After optimization, the separation system is considered.  The separation system determined is therefore case-specific, i.e., designed for a specific subset of mixtures.  Not all potentially feasible alternatives may have been identified and opportunities for an inherently cleaner process could be missed (Pennington 2000, 1999, 1997).

As a modification to such step-wise or hierarchical techniques, each system of a continuous process (reaction and separation) can be considered independently to identify alternatives.  Interactions between the two systems are then taken into account to further reduce the number of options requiring consideration and to determine potentially feasible flow diagrams for the overall process.  A subset or cluster of inherently cleaner process options can then be selected for further evaluation using economic, safety and environmental criteria. (e.g. Hoffmann & Hungerbuehler 1999, Pennington et al. 2000, Pennington & Yue 1999).  With few exceptions (e.g. Koller et al. 1999), the selection of a specific subset of options generally involves more detailed design to provide quantitative estimates of emissions, energy consumption, costs, etc.  This can be readily achieved using commercial process design simulation tools.

The prototype methodology required to help initially identify potential separation alternatives, independently of the reaction system, is described in this paper.  The prototype methodology used to analyze reaction systems, the overall framework and environmental selection criterion are described elsewhere (e.g. Pennington 2000, 1999, 1997).  Following the overall design philosophy outlined above, the mixture compositions and concentrations associated with the exit stream of the reaction system are not pre-defined (e.g., the use of a reactant in excess to ensure a given or complete conversion of another is not pre-assumed).  All chemicals that may leave the reaction system are considered.  This broadens the initial search space, hence the number of options that can be taken into account, but also introduces some new problems.

The screening-level prototype presented in this paper is applicable for continuous chemical processes and for the separation of mixtures of liquids, vapors and gases.  The prototype is primarily based on adaptations of the existing synthesis methodologies of Jaksland (1995, 1996), Smith (1995) and Barnicki (1990, 1991, 1992).  The current capabilities of the prototype, its relationship with the existing methodologies and key limitations are described.  Implications of the alternatives for the reaction system on the final mixture compositions, concentration ranges, which of the splits addressed will be required and the ultimate separation sequences are illustrated elsewhere (Pennington 2000, 1999, 1997).

Conclusions
The reaction and separation systems of a continuous process can be considered independently to identify alternatives.  Interactions between the two systems are then taken into account to reduce the number of alternatives and to determine potentially feasible flow diagram options for the overall process.  A subset or cluster of inherently cleaner processes can then be selected for further evaluation using economic, safety and environmental criteria.

The prototype required to help initially identify separation options for a process independently of the reaction system is described in this paper.  This prototype is applicable for the separation of liquids, vapors and gases in continuous chemical processes.  A number of separation techniques are considered.  Alternatives are not determined for case-specific mixtures:

• all components that may be in the stream leaving the reaction system of a process are considered,
• all potential splits, which may be required, are considered and
• case-specific concentrations are not assumed.

After consideration of the reaction and separation system interactions, the splits, techniques and sequences requiring further evaluation will be subsets of the separation options identified using the presented prototype.

Further research is required to:
 

• expand and evaluate the preference hierarchies of separation techniques from a pollution prevention (P2) perspective (including the assumed preference of separation techniques like phase splitting and simple distillation),
• evaluate the screening heuristics used to identify potentially feasible separation techniques for binary splits, 
• account for simultaneous exploitation of differences in more than one property,
• automate the identification of potentially feasible operating condition (concentrations, temperature, pressure) ranges for each split/technique and for each set of potential mixture compositions (identified from the reaction system analysis),
• account for azeotropes, liquid-liquid immiscibility, etc. for each set of potential mixture compositions (see Barniki 1991 and Jaksland 1996), and

automate the identification of potential separation sequence alternatives for each reaction system option for further evaluation.