Engineering decisions based on a narrow or reductionist point of view are often economically suboptimal and may not be environmentally sustainable. Such decisions contribute to increased exposure to risk, vulnerability to unforeseen events, loss of profitability, limited ability to identify new business opportunities, and ecological deterioration. Process systems engineering (PSE) adopts a more holistic view that considers the system of interest such as a process or product, along with its supply chain, goods and services that support the system. Such a view has led to many advances in engineering tasks such as design, control, and manufacturing. With the urgent need to improve the sustainability of human activities, the boundary of PSE is expanding to consider the entire life cycle of technological activities.
Our research is motivated by these challenges and is at the interface of multiple disciplines including engineering, ecology and economics. Most of our research projects aim to understand and enhance the sustainability and efficiency of existing and emerging technological systems. Typical projects include the following.
Life Cycle Oriented Methods
Making environmentally sound decisions requires a life
cycle view that goes beyond the process boundary and includes the
contribution of other industrial and ecological systems that provide
direct and indirect support to the selected activity. Such methods
have been popular but continue to present many challenges and
opportunities for research. Our work is developing new theory,
methods, tools and applications in this emerging field.
-
Accounting for Ecosystem Services – Since the sustainability of all human activities relies on ecosystem services, it is essential to account for them in life cycle methods. We are developing new methods to include the contribution of ecosystem services and identify vulnerabilities and risks due to ecosystem deterioration. Software based on an integrated economic-ecological model of the U.S. economy is available at http://resilience.osu.edu/ecolca.
-
Theoretical Framework – Life cycle oriented methods face challenges due to uncertainties and errors in inventory data, and difficulty in drawing a system boundary due to availability of models and data at multiple scales such as equipment, industrial processes and economy. We are introducing greater rigor in life cycle methods by imposing the laws of thermodynamics via a statistical framework.
-
Applications – Our methods are applied to a large variety of existing and emerging technologies including transportation fuels, nanomanufacturing, green chemistry, etc.
Designing Resilient and Sustainable Systems
This research strives to understand the characteristics of resilient and sustainable systems, for better management and design. Due to the difficulties in identifying and quantifying sustainability, or focus is on maintaining the resilience of a system to perturbations due to turbulence in economic, ecological, and societal factors.
-
Toward Self-Reliant Technological-Ecological Networks – The underlying premise of this work is that technological systems should be designed along with their supporting ecosystem services. Then the resulting Technological-Ecological Networks (Teco-Net) can be self-reliant. This work connects industrial ecology with ecosystem ecology and includes the role of ecosystems such as forests, wetlands, etc. in engineering decision making.
-
Complex Network Analysis and Modeling – Networks of technological and ecological systems are inherently complex with emergent properties that cannot be predicted based on reductionist models of individual systems. We are using methods such as network analysis and agent-based modeling to gain insight into such systems for enhancing their resilience and sustainability.
-
Process Design and Operation – Faced with increasing constraints on resources such as fossil fuels and water, industrial processes need to be optimized to minimize their use. We are developing methods based on thermodynamic concepts to enhance the efficiency of individual processes and their supply chain. We integrate process models with life cycle and economic models to understand the impact of policies such as carbon trading or taxes and environmental regulations.
Data Analysis and Modeling
Given the ease of obtaining measured data about chemical processes it is essential to have methods that can extract useful information from such data. In addition to measurements, models based on fundamental understanding and heuristic knowledge are also available.
-
Multiscale Analysis and Modeling - We have pioneered the development and use of multiscale data analysis methods using signal processing techniques such as wavelet analysis. New methods have been developed for multivariate Statistical Quality Control and nonlinear empirical modeling.
-
Bayesian Modeling - We are using Bayesian statistics to combine all kinds of information, including measurements, models and heuristics, for decision making. This work is relevant to solving tasks such as fault detection and diagnosis, system identification, linear and nonlinear estimation and optimization, and predictive control.
-
Statistical Analysis of Life Cycle Inventories – Since life cycle methods can be highly data intensive, models that can estimate life cycle impact based on variables that are easier to obtain are appealing. Such models are also attractive for preliminary screening of emerging technologies and at early stages of decision making.