(246c) Superstructure-Based Industrial Wastewater Treatment System Synthesis

Stricter environmental effluent standards have forced industries to target increased water-use efficiency and water reuse.  In order to meet this demand, systematic methods based on superstructure optimization for WN synthesis have been developed.  Mathematical programming approaches are taken to optimize superstructures to generate the optimal water network for the system using NLP or MINLP models^[2,6].  Water is supplied to water-using process units, and then wastewater streams generated from these processes are treated in various treatment units.  In the typical superstructure-based models^[7], process units are defined by their allowable contaminant concentration level, whereas wastewater treatment units are usually defined based on their contaminant removal ratio.  This versatile superstructure considers systematic alternatives for water reuse, recycle, and recycle-reuse to minimize freshwater consumption or total network cost subject to discharge limit imposed.  However, these models greatly simplify wastewater treatment units by using fixed recoveries, creating a gap for their applicability to industrial processes.  It is clear that further improvements can be obtained by considering more accurate treatment models in the optimization of these water networks.

This presentation describes a unifying work combining various technologies capable of removing all the major types of contaminants through the use realistic models for the treatment units.  Improvements over the simplified model are as follows.  First, unit-specific short-cut models are developed in place of the fixed contaminant removal model to describe contaminant mass transfer in wastewater treatment units.  Only a few works in the literature have considered more detailed models of treatment processes and these approaches are limited to a single treatment technology ^[4,8].  In this work, multiple types of treatment units and appropriate modeling equations that can satisfactorily predict unit performance with reasonable computational complexity are considered.  For example, the rapid sand filter model and the Ergun equation are used to describe filtration units, and the osmotic pressure model is used to calculate the transmembrane pressure and the membrane area needed for reverse osmosis units.  For units that do not demand complex mass transfer correlations (e.g. settling tank, filter press), new removal efficiency relations are developed to account for the dependency of the removal ratios on the presence of other contaminants^[5]and the temperature of the unit for a more rigorous representation of these treatment units. 

Furthermore, the architecture of the superstructure is modified to accommodate realistic potential structures. Different types of contaminants (suspended solids, dissolved solids) present in the system are removed by considering the Best Available Technology (BAT)^[1], which provides the standards for discharge of priority pollutants.  Since there are multiple treatment technologies for the removal of each type of pollutants, the modified superstructure allows for the selection of a subset of BAT treatment technology through the use of disjunction in the MINLP formulation. Another drawback in the original WN superstructure is that freshwater supplied to the system is assumed to be of high purity and therefore can satisfy the contaminant concentration requirement at the inlet of the process units.  Yet this is not always the case in industrial applications, especially when freshwater is serving the purpose of feedwater to boilers.  This work accounts for raw water treatment subsystem, which is placed prior to the water-using subsystem and wastewater treatment subsystem to accommodate more interactions in the overall WN superstructure^[3]_.  Accordingly, the set of raw water pretreatment units can be used to process water streams with higher purity than the set of wastewater treatment units used to treat wastewater streams.

Finally, short-cut wastewater treatment cost functions (O&M cost and construction cost)^[9]in the form of polynomial functions are incorporated into the model.  The conventional network cost function usually consists of an operating cost term (linear with flowrate) and a capital cost term (exponential) for the sake of convenience and ease in computation.  The use of more complex objective in this more rigorous model enables the capability to design different WNs that allow for trade-offs that better meet the need of their respective decision criteria.  The resulting superstructure, which is modeled as an MINLP, is illustrated through several industrial applications in metal finishing, and pulp and paper industries.

References:

[1] Integrated pollution prevention and control reference document on available techniques in common waste water and waste gas treatment/management systems in the chemical sector, Tech. Rep., European Commission, BREF document, Current Draft, 2009.

[2] M. J. Bagajewicz, “A review of recent design procedures for water networks in refineries and process plants,” Computers & Chemical Engineering, vol. 24, no. 9-10, 2000, pp. 2093 – 2113.

[3] M. J. Bagajewicz and D. C. Faria, “On the appropriate architecture of the water/wastewater allocation problem in process plants,” Computer Aided Chemical Engineering, vol. 26, 2009, pp. 1-20.

[4] E. Bringas, R. Karuppiah, M. F., San Roman, I. Ortiz, and I.E. Grossmann, “Optimal groundwater remediation network design using selective membranes,” Ind. Eng. Chem. Res., vol. 46, 2007, pp. 5555-5569.

[5] B. Galan and I. E. Grossmann, “Optimal design of distributed wastewater treatment networks,” Industrial & Engineering Chemistry Research, vol. 37, no. 10, 1998, pp. 4036–4048.

[6] J. Jezowski, “Review of water network design methods with literature annotations,” Industrial & Engineering Chemistry Research, vol. 49, no. 10, 2010, pp. 4475–4516.

[7] R. Karuppiah and I. E. Grossmann, “Global optimization for the synthesis of integrated water systems in chemical processes,” Computers & Chemical Engineering, vol. 30, no. 4, 2006, pp. 650 – 673.

[8] Y. Lua, Y. Hua, X. Zhang, L. Wu, and Q. Liu, “Optimum design of reverse osmosis system under different feed concentration and product specification,” Journal of Membrane Science, vol. 287, 2007, pp. 219-229.

[9] J. Sharma, Development of a preliminary cost estimation method for water treatment plants, master’s thesis, University of Texas at Arlington, 2010.