Search

The observed induction time for seeded gypsum crystallization was investigated with respect to gypsum particle seed size and mass loading. The induction time was significantly shortened in the pres...

Sensor fault detection and isolation (SFDI) approaches, based on support vector regression (SVR) plant sensor models and self-organizing-map (SOM) analysis, were investigated for application to reverse osmosis (RO) desalination plant operation. SFDI-SVR and SFDI-SOM were assessed using operational data from a small spiral-wound RO pilot plant and synthetic faulty data generated as perturbations relative to normal plant operational data. SFDI-SVR was achieved without false negative (FN) detections for sensor deviations of ≳|10%| and FN detections of, at the most, ≲|5%|, and for sensor deviations of ≳|4%| at sensor fault detection (FD) thresholds of up to ∼|4%|. False positive (FP) detections were almost invariant, with respect to sensor FD, being ≲|5%| for sensor deviations of ≳|5%|. Corrections of faulty sensor readings were within SVR model accuracy (AARE < 1%) for SFDI-SVR and ≲|5%| for SFDI-SOM. Although SFDI-SOM has lower detection accuracy, it requires a single overall plant model (or SOM), while providing pictorial representation of plant operation and depiction of faulty operational trajectories.

A novel model-based control system for the operation of a spiral-wound RO membrane desalination system was developed with a focus on maintaining energy-optimal operation. The control scheme utilized an operational model for spiral-wound RO desalting with a supervisory controller providing real-time updates of membrane permeability and the appropriate feed pressure set-points for maintaining the target permeate productivity at the lowest feasible specific energy consumption. System RO feed pressure and flow rates were controlled by a lower-level RO controller through adjustment of the RO high pressure feed pump, variable frequency drive, and RO concentrate valve. Seawater desalination tests with an RO plant, capable of permeate productivity up to 18 000 gallons/day, demonstrated effective self-adaptive energy-optimal operation, subject to feed salinity fluctuations, constraints imposed by the system’s physical limitations (i.e., minimum and maximum feasible operational pressures and flow rates), and the th...

This work focuses on the design and implementation of a nonlinear model-based control system on an experimental reverse-osmosis (RO) membrane water desalination system to address large set-point changes and variations in feedwater salinity. A dynamic nonlinear lumped-parameter model is derived using first-principles, and its parameters are computed from experimental data to minimize the error between model predictions and experimental RO system response. This model then is used as the basis for the design of a nonlinear control system, using geometric control techniques. The nonlinear control system is implemented on the experimental RO system, and its set-point tracking and disturbance rejection capabilities are successfully evaluated.

The potential impact of coagulants on mineral scaling in reverse osmosis (RO) feed treatment of brackish water was assessed experimentally, with respect to calcium sulfate and barium sulfate scalin...

Advances in highly permeable reverse osmosis (RO) membranes have enabled desalting operations, in which it is practically feasible for the applied pressure to approach the osmotic pressure of the exit brine stream. However, energy cost remains a major contributor to the total cost of water produced by RO membrane desalination. Reduction of the overall cost of water production represents a major challenge and, in the present work, various elements of water production cost are evaluated from the viewpoint of optimization, with respect to various costs (energy, membrane area and permeability, brine management, and pressure drop), as well as the important thermodynamic cross-flow constraint, utilization of energy recovery devices, and operational feed and permeate flow rate constraints. More specifically, in the present study, an approach to the optimization of product water recovery at pressures that approach the osmotic pressure of the exit brine stream is presented via several simple RO process models that...

Model-based control and monitoring such as feed-forward/feedback control, fault detection and isolation (FDI), and fault-tolerant control (FTC) techniques that utilize Lyapunov-based control laws are implemented on a high recovery reverse osmosis desalination plant model. A detailed mathematical model of a high recovery reverse osmosis plant is developed. This model incorporates the large spatial variations of concentration and flow rate that occur in membrane units during high recovery operation. Bounded nonlinear feedback and feed-forward controllers are developed and applied to this system. The application of these controllers with FDI and FTC is demonstrated in the context of a high recovery reverse osmosis process simulation. The first set of simulations demonstrates the ability to compensate for the effects of large time-varying disturbances in the feed concentration on specific process outputs with and without feed-forward control. The second set of simulations demonstrates the ability of FDI and F...

A modified expression for the change in chemical potential of a solvent, in a polymer network, due to isotropic swelling was obtained by substituting a non-Gaussian chain length probability distribution in Flory's statistical analysis of rubber elasticity. The affine non-Gaussian expression for the free energy change due to elastic deformation was compared to both the traditional Flory and the James-Guth phantom network expressions for the free energy change due to elastic deformation using available experimental data for the swelling of highly cross-linked polymeric resins. Among the different models for the elastic contribution to the free energy change resulting from isotropic swelling, the non-Gaussian elasticity expression was found to be as good as, or slightly better than the Flory expression, and clearly superior to the James-Guth phantom network expression. Both of the two different expressions used to represent the free energy of mixing, the Flory-Huggins expression and a modified version of the quasi-chemical mixing expression, were found to be equally successful when describing the isotropic swelling of cross-linked polymeric resins

A screening-level methodology was developed for the evaluation of solute affinity for polymers that are candidate sorption resins. In this approach novel grafted polymer-silica resins were synthesized to produce poly(vinylpyrrolidone)-silica (PVP-Si) and poly(vinyl acetate)-silica (PVAc-Si) resins. The polymer-silica resins along with a number of commercially available polymer resins were used to evaluate the aqueous-phase adsorption of phenol, tetrachloroethene (PCE), trichloroethene (TCE), and chloroform. The polymer-grafted silicas were able to selectively remove pollutants from water with a covalently bonded polymer layer that has a high affinity for the target pollutant. The PVAc-silica resin had a sorption capacity for TCE and ChCl 3 as high as commercial poly(styrene) resin XAD-4; the PVP-Si resin had a sorption capacity for phenol higher than a commercial poly(vinylpyridine) resin (Reillex 425). PCE adsorption onto the PVAc-silica was comparable to the commercial poly(methacrylate) and poly(vinylpyridine) resins but less than the poly(styrene) resin. The results show that the Hildebrand solubility parameter along with the dipole moment of the polymer functional groups can be used for an initial screening-level assessment of polymer-solute affinity

We have extend the preliminary study of Chaimberg et al. (1989) on the graft copolymerization of vinylpyrrolidone to a wider range of reaction temperatures and initial monomer concentrations. In addition, we report on an experimental procedure for detaching the graft polymerized PVP chains from the substrate surface. The average molecular weigth of the degrafted polymer was determined by way of HPLC-SEC using a novel PVP-silica matrix separation material.

A new computational model for free-radical polymerization reactions was developed. The model algorithm is based on the use of an implicit numerical technique to solve for the coupled monomer and total growing polymer concentration differential equations. Determination of the monomer and total growing polymer concentrations ensures that conservation of mass is maintained without the need to solve the large number of differential equations associated with the growing polymer concentrations for each individual chain length. Acceleration of the numerical calculations was achieved by lumping the growing polymer chains into distinct groups. The proposed numerical scheme generates the monomer conversion and a complete molecular weight distribution (MWD) of the growing and dead polymer chains. The current formulation allows for the incorporation of nonidealities, such as the onset of gelation and chain-length-dependent reaction rate coefficients

The adsorption of the EPA priority pollutants trichloroethene (TCE) and chloroform (CHCl 3 ) from aqueous solutions onto activated carbon and macroporous polymeric resins was investigated over a wide concentration range. Over much of the concentration range studied, activated carbon adsorbed more TCE and CHCl 3 on a weight basis and on a surface area basis than did the polymer resins. At concentrations greater than 1000 μg/L, the adsorption capacity, on a surface area basis, for the different resins and for activated carbon was similar. The adsorption isotherms were fitted by the empirical Freundlich isotherm and the semiempirical isotherm of Jossens et al. The affinity of the solutes for the different resins was qualitatively described by a simple Hildebrand solubility parameter correlation