• Home
• Search
• About
• Contact Us
• Site Map

• News & Events
• Articles
• Tools
• Links
• Virtual Plant Tour

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Consider the following recent examples of pollution prevention in chemical process plants:

• An Adiponitrile plant in Orange, Texas installed an inexpensive solenoid valve on the line supplying process water to flush the seals of process pumps used in intermittent service. The solenoid, which is connected to the pump control system, shuts off flush water when the pumps are not running. As a result, the plant has reduced its waste water volume by 57 million gallons per year.

• The same facility upgraded the catalyst decanting system in a petrochemical process, permitting higher recovery rates of catalyst and higher yield in the process. Solid wastes were reduced by approximately 1,000,000 pounds per year, and the investment in new equipment paid for itself in approximately 18 months.

• As part of an overall effort that reduced waste generation by 98%, Dow Corning installed mechanical separators between a fluid bed reactor and a distillation column in a silane manufacture process. The separators reduced the amount of solids which had been settling in the column. The process modification reduced the frequency of maintenance and cut the amount of product lost when solids are pumped out of the column for disposal.

Examples of cost-effective process modifications such as these are becoming increasingly commonplace as pollution prevention becomes a more important component of the chemical industry’s environmental management philosophy. Indeed, through industry initiatives such as Responsible Care™ and the recently adopted ISO 14000 environmental management standards, pollution prevention has become institutionalized as a "mainstream" environmental management practice.

Given the success of pollution prevention at the plant level, it is not surprising that the industry is increasingly looking to incorporate the concept in the earliest stages of process and product development -- that is, in process and product design.

The integration of environmental considerations in design and optimization is commonly referred to as "Design for Environment," or DfE. And it seems to be catching on. A recent review of Environmental Annual Reports (EAR’s) published by leading chemical companies highlighted this movement towards a design for environment perspective as one of the key trends in environmental management practices. And while chemical engineers have not developed as comprehensive a view of DfE as one might encounter in the design of manufactured goods a number of articles have been published which can help process design engineers apply DfE to process design.

The Role of Heuristic Methods in Design

Applying pollution prevention to the design of new processes presents some interesting technical and organizational challenges. Process design engineers must often make preliminary design decisions on the basis of incomplete information and untested assumptions. As design proceeds and investment in the process is made, these decisions develop enormous "momentum" of their own, making it increasingly difficult to revise the choices. Compared to product design, where early prototyping can permit the parallel evaluation of several design options, process design "locks in" choices earlier and with more force, putting even more of a premium on making good choices early on.

In many aspects of process and product design, engineers rely heavily on the use of design heuristics. In the classic engineering design text by Pahl and Beitz, a heuristic is described as "explicit knowledge [and] non-explicit knowledge…necessary in order to organize the sequence of thinking operations, including modifying operations (searching and finding) and testing operations (checking and assessing)." More commonly, a heuristic is a general procedure or rule of thumb, which is used to suggest solutions or strategies for solving a problem, often in the absence of "deep" knowledge about a system.

Central to the concept of heuristic design is the notion that a heuristic is not necessarily prescriptive solution. More commonly, a heuristic simply suggests a method of attack for solving a particular problem. For example, office equipment manufacturers such as AT&T and Xerox are increasingly concerned about producing products which can be easily recycled. One general approach to increasing the recyclability of a product is to standardize key durable components so that the components can be reused, or the equipment more easily remanufactured. Thus, "enforce standardization in the design of durable components" (or, more simply, "design for re-use") may be thought of as a heuristic for environmentally-sound product design. In the design of electronic equipment, automobiles, and other manufactured products, considerable effort has been directed at articulating a set of fundamental DfE design heuristics. A more complete listing of design heuristics for manufactured products is listed below in Table 1.

Table 1. Some common Design for Environment ("DfE") heuristics applicable to product design. (Adapted from Fiksel). 

Unfortunately, a comparable set of design heuristics for process designers does not yet exist. While a number of guidelines for incorporating pollution prevention in process design have been published, the process-specific nature of the recommendations and the lack of an organizing framework limits their effectiveness or limits their range of application.

Process Heuristic Review for Environmental Design (PHRED)

Recent work at Battelle has attempted to address this gap through the development of a framework known as Process Heuristic Review for Environmental Design, or PHRED. PHRED is intended as a design review process which can be used at all stages of process design to provide guidance regarding conceptual design, equipment selection, equipment specification, control system and piping design, and even the design of ancillary facilities such as stormwater collection systems and control rooms.

Just as the current crop of pollution prevention success stories resulted in part from the development of effective pollution prevention opportunity assessment methods, the routine and effective integration of pollution prevention into process design will depend in great part on the development of new tools which adapt these techniques to the needs of the designer. In developing PHRED, we have attempted to respond to some of the characteristics of process design:

• The process designer, especially in early stages of design, must contend with incomplete or inaccurate information about the process. If a design review process is to be useful, it must be able to provide meaningful guidance even in situations where physical properties, kinetic data, and details of the process chemistry are unknown and/or uncertain.

• Design proceeds through several more or less distinct phases, each of which requires review guidance unique to that phase. For example, process design often starts with a conceptual design phase, during which basic unit operation requirements are established, and overall material balances are developed. This is generally followed by an equipment selection phase in which the functional requirements of the major unit operations are translated to preliminary design of major process equipment, in which operating pressures, temperatures, and related parameters are determined. Subsequently, detailed equipment specification establishes the rigorous optimization, selection of materials of construction, and detailed mechanical design of process vessels. Finally, design of piping and control systems must be completed, often concurrently with the detailed design. Of course, other classifications exist for design phases, but the important point is that at each phase, the focus of design recommendations provided by a review must be appropriate to the task both in terms of their prescriptive content and their information requirements.

• Design processes, especially in equipment specification, are increasingly being automated via the application of tools such as dynamic process simulation and optimization. Consequently, it is critical to establish linkages between pollution prevention design tools and the software applications used by design engineers.

Consistent with its name, PHRED is being developed as a design review process, albeit an automated one. Although the framework developed for PHRED can be applied without the use of software, the ultimate objective of this effort is to develop a knowledge-based software application which can be used by designers to identify and prioritize applicable environmental design strategies. These strategies are designed to be used in conjunction with more rigorous design tools including lifecycle inventory analysis, simulation-based optimization and structured decision analysis tools such as the Analytical Hierarchy Procedure (AHP).

In developing PHRED, we have attempted to adhere to some basic concepts:

• Recommendations derived from PHRED should be based in practical experience of process plant operations, applying to existing technology and current design practice wherever possible. While state of the art technology may play a role in pollution prevention, it is important to recognize the need to provide designers with practical solutions to common design problems.

• While process plant experience is an important starting point, it is critical that the design principles in PHRED also be rooted in fundamental process phenomena (e.g., kinetics, thermodynamics, and fluid dynamics). This permits the lessons of the process plant to be applied under an extended range of conditions, including in novel chemistries and processing environments where case study information rarely exists.

• The value of heuristics can be enhanced by representing them as a hierarchical structure, starting with a core set of general guidelines which embody a larger number of more specific design strategies. As Pahl and Beitz point out, there are important cognitive reasons for this: structuring the heuristics in this manner makes it easier for a designer to remember the concepts, and allows the designer to make incremental decisions about which strategy(ies) to apply rather than having to select strategies from the entire set.

• Where possible, the design strategies in PHRED should suggest more rigorous simulation and optimization methods, so that the respective strengths of both heuristic and formal analytical methods are applied to the design problem.

• The heuristic should reflect "hidden" wastes which may not appear on process flowsheets. Many wastes in process plants, especially those found in batch plants, occur during process transients, as a result of maintenance, or as fugitive emissions due to mechanical loss. It is important that these wastes be incorporated in the evaluation of design options.

Since one of the underlying assumptions in developing PHRED is that practical plant experience should serve as the basis for devising design strategies, our begins in the process plant. Starting from a collection of more than two hundred case studies and examples of pollution prevention in the process industries, we identified dozens of examples of process and equipment modifications which are applicable to process design. From these case studies, a series of candidate design principles, or heuristics, were developed. These were supplemented by design guidance provided in articles such as the excellent series by Nelson.

Although the design strategies and heuristics included in PHRED are not required to meet all of the above criteria, these guidelines were very helpful in selecting and identifying candidate strategies. The candidate strategies were then reviewed to identify common themes, concepts, and approaches and where possible, similar concepts are combined. The combined strategies are included within the PHRED framework as high level design heuristics, applicable to a wide range of processes and conditions. Useful design concepts which did not meet the above criteria, but which appear to have validity in a range of process applications, were incorporated at the lower levels of the PHRED framework. Viewed as a tree, heuristics appear at the root of the tree, more specific strategies in the branches, and process or equipment-specific strategies at the ends of the branches. An illustration of this approach is shown in Figure 1.

Figure 1. "Tree" representation of a structured heuristic. Note that heuristics begin with broad goals and branch out to increasing levels of specificity.

Results to Date

Our work to date has resulted in the tentative identification of four high-level heuristics and a number of more specific design strategies. While we are continuing to develop these heuristics and to incorporate them into PHRED, the following appear to be key concepts:

• Design to reduce maintenance wastes. Maintenance-related wastes include solvents used for vessel cleaning, process fluids drained during servicing of pumps and heat exchangers, cleaning agents and abrasive materials used in cleaning heat exchange tubes, and heavy-phase materials such as tank heels, suspended solids, and sludge which must be removed due to gradual accumulation in process vessels. Design strategies include providing interim storage for process and cleaning fluids, minimizing the volume of piping runs which must be drained for maintenance, reducing maintenance frequency by using foul-resistant coatings and materials, and insuring that frequently-cleaned vessels are provided with adequate drainage and line-of-sight clearance for high-pressure washing.

• Design for selectivity. Another common source of wastes are byproducts created incidentally in separations equipment, heat exchangers, and other non-reactor vessels. Many of these operations result in the formation of byproducts from thermal decomposition or incidental reaction of process fluids. Design strategies include reduction of reboiler temperatures by elimination of process "hot spots," use of reduced distillation column pressures or specification of polymer-lined or high-polish finishes in heated vessels.

• Design to reduce mechanical losses. Mechanical losses of materials include fugitive emissions due to tank vents, pump seals, and valve packing; they can also include losses due to steam-injection vacuum pumps, sampling loop purges, and spillage from dry solids transfer operations. Design strategies include specification of vent recovery systems, specification of sealless pumps or closed-circuit seal rinses, and reducing valve and flange counts.

• Design to reduce transient wastes. Transient wastes occur during process upsets and start-up/shut-down operations. Design strategies include providing interim storage and recycle loops and tankage for recovering off-spec product, implementing parallel reactor systems, and improving process control techniques.

During the classification of design strategies and heuristics, it became obvious that not all of the strategies identified thus far can be applied equally well during all phases of design. Additionally, while some strategies apply well to reactor design, others are more aptly applied in separation processes. Table 2 provides more specific illustrations of DfE strategies suggested by the process design heuristics, and indicates which phases of process design they are best applied.

Table 2. Applicability of process design heuristics during different stages of process design.

Where is PHRED Headed?

Taken by themselves, the heuristics and design guidance identified through our work can be applied to a wide range of processes and process conditions. In their present form, they serve as a basis for developing checklist approaches to design review and approval processes. In fact, this approach has been successfully applied by co-workers for the US Department of Energy in the area of facility design. We also expect to expand the list of heuristics based on further analysis of pollution prevention case studies.

However, the our experience has shown that as the list of heuristics grows, it becomes increasingly important to provide a mechanism for identifying the most appropriate strategies and prioritizing amongst them. It is also important, in light of the increasing role of software in engineering design, to establish linkages between PHRED and the simulation and design applications used by process engineers. Accordingly, much of our current activity is focused on determining which factors determine the relative utility of individual design strategies, and placing this knowledge in the framework of a rule-based expert system. These distinctions between design strategies, which are based on the scope of design issues being considered, types of equipment being examines, range of process conditions, etc. are currently being used to develop an expert system which will help designers select the most appropriate strategies for a given process.

Conclusions

The primary reason for incorporating pollution prevention into process development and design is one of cost effectiveness. Decisions made early in the development process often determine later development activities including laboratory and pilot plant studies, equipment and materials selection, and project economic analysis. By addressing environmental issues early in the development cycle, unforeseen technical, regulatory, and economic consequences of design choices can be anticipated. The net result is a reduction in the technical and economic risk associated with environmental issues. Additional benefits may also be obtained when concurrent consideration of environmental issues with other engineering factors leads to quicker time-to-market, process innovation, improved quality of products, or increased efficiency.

Development of improved tools and methods for integrating pollution prevention into the design of new processes is a promising application for information technologies ranging from robust simulation and optimization to artificial intelligence and decision support tools. Through the development and application of these tools, we can expect success stories such as those at the beginning of this article to become more commonplace.

Acknowledgments

Much of the conceptual approach described in this article has been developed under a collaboration between the US EPA, the US Department of Energy, and the Center for Clean and Industrial Treatment Technologies (CenCITT), with participation from the Center for Waste Reduction Technologies (CWRT). The author would like to acknowledge the support and involvement of these organizations in the preparation of this paper.