Recycle/reuse networks are commonly used in industrial facilities to conserve natural resources, reduce environmental impact, and improve process economics. The design of these networks is a challenging task because of the numerous possibilities of assigning stream (process sources) to units that may potentially employ them (process sinks). Additionally, several fresh streams with different qualities and costs may be used to supplement the recycle of process streams. The selection of the type and flow of these fresh resources is an important step in the design of the recycle/reuse networks. This work introduces systematic approaches to address two new categories in the design of recycle/reuse networks: (a) The incorporation of thermal effects in the network. Two new aspects are introduced: heat of mixing of process sources and temperature constraints imposed on the feed to the process sinks iv (b) Dealing with variation in process sources. Two types of source variability are addressed: flowrate and composition For networks with thermal effects, an assignment optimization formulation is developed. Depending on the functional form of the heat of mixing, the formulation may be a linear or a nonlinear program. The solution of this program provides optimum flowrates of the fresh streams as well as the segregation, mixing, and allocation of the process sources to sinks. For networks with variable sources, a computer code is developed to solve the problem. It is based on discretizing the search space and using the concept of "floating pinch" to insure solution feasibility and optimal targets. Case studies are solved to illustrate the applicability of the new approaches.

Over 18 billion barrels of waste fluids are generated annually from oil and gas production in the United States. As a large amount of water is used for oilfield operations, treating and reusing produced water can cut the consumption of fresh water in well sites. This research has helped to develop a membrane process train for a mobile produced water treatment unit for treating oilfield produced brine for reuse. To design the process train, over 30 sets of combination tests at pilot laboratory scale were performed using pretreatment, microfiltration and nanofiltration processes. Membrane performance was selected based on high flux separation efficiency, high tolerance for solids and fluid treatments. Over 95 % solids rejection and greater than 80 % oil removal efficiency were obtained in all these tests. Process train (pre-treatment and membrane) performance was monitored by chemical analysis of permeate and models fitting experimental data for the process. From the results, hydrocarbon rejection was analyzed; total organic carbon rejection was 47.9 %, total carbon content averaged 37.3 % rejection and total inorganic carbon rejection was at 3.66 %. BTEX removal efficiency ranged from 0.98 % to 52.7 % with the progressive pretreatment methods of using cartridge filters. The nanofiltration membrane showed significant reduction in total dissolved solids and in both anionic and cationic species. The process train is seen to follow a sequence of treatment from cartridge and oil removal filter treatment to microfiltration treatment to ultrafiltration, followed by nanofiltration for the purpose of this research. Further research still needs to be done on to determine the kind of analytical test which will give real time feedback on effectiveness of filters. In summary, the process train developed by TAMU-GPRI possesses distinct advantages in treating oilfield produced brine using membrane technology. These advantages include high quality of permeate, reduced sludge and the possibility of total recycle water systems. The small space requirement, moderate capital costs and ease of operation associated with the use of the mobile unit membrane technology also makes it a very competitive alternative to conventional technologies.