Soft solutions to hard problems

In the information age, it is only fitting that significant research effort is being put into using intelligent computational techniques to address the world’s growing environmental problems. The use of mathematical or computer models to simulate the interactions of technology with the environment sometimes yields counterintuitive solutions not immediately evident from a layman’s perspective. For example, Kreith’s recent article in the Journal of Energy Resources Technology dispels many of the myths of the much-touted hydrogen economy envisioned by today’s energy specialists as the foundation of energy infrastructure for transportation in the future. Using straightforward life cycle-based thermodynamic arguments, the paper concludes that many of the proposed fuel hydrogen production and delivery systems, particularly those that rely indirectly on fossil fuel feedstocks, are actually less efficient – and consequently more environmentally damaging – than conventional systems in use today. Many similar environmental issues can be addressed through the use of mathematical modeling.

Closer to home, our research at the Sustainable Technology Laboratory in De La Salle University-Manila over the past decade has led to increased specialization in two specific areas – life cycle assessment (LCA) and pinch analysis. LCA methodology allows for a thorough accounting of environmental impacts associated with technologies, including indirect ones. For example, cellphones are not often regarded as generating air pollutants. A closer look reveals that the power used to charge cellphone batteries is often derived from a mix of electricity generated by different plants, including some which use coal, oil or gas as fuel. Thus, from an LCA perspective, cellphones generate greenhouse gases and other pollutants through their allocated share of power plant emissions. The actual procedure for computing this share, and accounting for it over the entire life of the phone, is somewhat more complicated. Considerable effort is being put into developing and perfecting such computational techniques. Techniques derived from operations research (OR) and artificial intelligence (AI) have been tested on different problem domains in recent years. In La Salle, we have made use of expert systems, possibility theory, rough and fuzzy sets, Monte Carlo, and linear programming in LCA. Applications have included assessment and comparison of alternative fuels for road vehicles, and the development of pollution prevention strategies for the production of bulk goods (such as pulp and paper) and manufactured products (such as semiconductor devices). Similar novel approaches are being developed by our counterparts elsewhere in the world. A good example is a recent NATO-funded joint project by the University of Portsmouth in the UK and Moscow State University that used a mathematical model to develop optimal and economically viable pollution prevention measures for the production of bricks. We are now mobilizing a collaborative effort to develop more efficient solution algorithms for the same design problem. Potential approaches identified thus far include genetic algorithms, fuzzy integer programming and inductive reasoning based on rough sets.

Mathematical models are intended to provide humans with simplified versions of reality to facilitate understanding of its behavior. It stands to reason that some resolution is lost in the process of translating real systems into a set of mathematical expressions or a computer code. The words of Zadeh, who developed the concept of fuzzy sets in the 1960s, summarize this dilemma:

"As the complexity of a system increases, our ability to make precise and yet significant statements about its behavior diminishes until a threshold is reached beyond which precision and significance (or relevance) become almost mutually exclusive characteristics."

Many of the tools we use are thus designed to deal with uncertainties inherent in the modeling process. In particular, we have been partial to the use of fuzzy numbers and fuzzy arithmetic, which offer some advantages with respect to computational techniques compared to conventional probabilistic approaches.

Pinch analysis is a highly specialized field that makes use of various mathematical techniques for efficient use and recycling of resources (such as energy or water) in manufacturing facilities. In the late 1970s and 1980s, much of the work in this field focused on energy conservation through heat recovery networks. The work of El-Halwagi since the late 1980s and the process integration group in UMIST in the early 1990s led to extensions of these techniques to conservation of industrial solvents and water. Current basic research in pinch analysis is focused predominantly on developing simple, computationally efficient procedures for finding optimal water budgets or "targets" for plants, as well as developing feasible recycle networks to meet such budgets. Two schools of thought have emerged. The first advocates the use of simple graphical displays with visual and intuitive appeal. An example is the water cascade technique developed a few years ago in Universiti Teknologi Malaysia (UTM). The other school of thought prefers an approach based on mathematical programming "superstructures," which allows a whole array of algorithms to be brought to bear on problems with different features. While criticized as a "black box" approach that yields little in the way of insight in developing these solutions, they are able to solve problems that would prove intractable using graphical methods. In La Salle, we subscribe to this second school of thought. We have also successfully demonstrated the first ever (that we know of) combination of pinch analysis with LCA concepts, thereby paving the way for the potential full integration of these two fields in the near future. That work was done as part of the master’s thesis of Valarie Ku-Pineda, and will be published in the near future in the Journal of Cleaner Production.

We have also put considerable effort into dissemination of these ideas to industry, government agencies, our colleagues from the academe and, of course, our students. Dedicated LCA courses for engineering have been taught at the graduate level in La Salle since the late 1990s. Some related concepts have also been integrated into a number of undergraduate courses as well as graduate programs in science and business. When LCA was offered at the masteral level earlier this year, a project-based approach emphasizing the principle of "learning by doing" was used with very encouraging results. At the moment a one-semester course on pinch analysis is also being offered in the master’s program using a similar approach. Seminars and workshops are other venues for disseminating these concepts to a wider audience. Through the Center for Engineering and Sustainable Development Research (CESDR) La Salle has hosted a number of these events. In the first quarter of 2005, the seminars on LCA and pinch analysis featured respected experts from the National Institute for Advanced Industrial Science and Technology (AIST) in Japan and UTM in Malaysia. The seminar participants included practicing engineers and managers from industry as well as officials of different government agencies and NGOs. In spite of these mild successes, much remains to be done; we are currently trying to develop innovative concepts for the delivery of these ideas, for example, through the Internet or through properly designed distance-learning materials.

As it happens, this area of research is also one that does not necessarily require extensive financial resources. This is a significant bonus for researchers and students working in the cash-strapped scientific environment of the Philippines. At the Sustainable Technology Laboratory, we have managed to produce internationally competitive work with only a handful of PCs, some freeware, and an ever-growing pool of talented research students. One major area of study by the research group is the assessment of alternative fuels and energy sources – which will be the subject of our next column.

* * *

Dr. Alvin B. Culaba is a professor of the Mechanical Engineering Department of De La Salle University-Manila. He is the director of the Center for Engineering and Sustainable Development Research (CESDR) and the head of the Sustainable Technology Laboratory.

Dr. Raymond R. Tan is an associate professor of the Chemical Engineering Department and the director of the Engineering Graduate School of De La Salle University-Manila.