Dealing with Uncertainties Arising from Environmental Conscious Multi-Objective Optimization
Process optimisation has been at the core of design and retrofit of process industries. Traditional process design focuses on plant operations to minimise costs or maximise profits using performance indicators of conversion, yields, efficiency, and productivity. With dwindling resources coupled with...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Conference Paper |
| Published: |
Elsevier BV
2012
|
| Online Access: | http://booksite.elsevier.com/9780444594310/papers.php http://hdl.handle.net/20.500.11937/29772 |
| Summary: | Process optimisation has been at the core of design and retrofit of process industries. Traditional process design focuses on plant operations to minimise costs or maximise profits using performance indicators of conversion, yields, efficiency, and productivity. With dwindling resources coupled with environmental degradation, environmental objectives started to be incorporated into the optimisation, through strategies such as waste minimisation and pollution prevention. While environmental conscious design has benefited from life cycle assessment in finding the environmental burdens of a process, it generates a multi-objective optimisation problem with many variables and constraints. Mathematical programming provides a solution in the form of Pareto front, though to find a compromised solution remains a difficult task. Furthermore, the incorporation of environmental indicators brings in a considerable number of uncertainty sources that exacerbate the difficulties of optimization. These reflect both on economic costs and environmental conditions. In this paper, the progress and framework of incorporating LCA results with process design and optimisation is given followed by a discussion of barriers, with the emphases on uncertainty issue. A case study on enhanced oil recovery from a reservoir using CO2 is carried out to highlight the effect of uncertainties in one of the external index to the optimisation results. |
|---|