skip to main content

Explore Research Products in the NSF-PAR

Title: A Neural Network Based Superstructure Optimization Approach to Reverse Osmosis Desalination Plants
An ever-growing population together with globally depleting water resources pose immense stresses for water supply systems. Desalination technologies can reduce these stresses by generating fresh water from saline water sources. Reverse osmosis (RO), as the industry leading desalination technology, typically involves a complex network of membrane modules that separate unwanted particles from water. The optimal design and operation of these complex RO systems can be computationally expensive. In this work, we present a modeling and optimization strategy for addressing the optimal operation of an industrial-scale RO plant. We employ a feed-forward artificial neural network (ANN) surrogate modeling representation with rectified linear units as activation functions to capture the membrane behavior accurately. Several ANN set-ups and surrogate models are presented and evaluated, based on collected data from the H2Oaks RO desalination plant in South-Central Texas. The developed ANN is then transformed into a mixed-integer linear programming formulation for the purpose of minimizing energy consumption while maximizing water utilization. Trade-offs between the two competing objectives are visualized in a Pareto front, where indirect savings can be uncovered by comparing energy consumption for an array of water recoveries and feed flows.  more » « less
Award ID(s):
1739977
NSF-PAR ID:
10463739
Author(s) / Creator(s):
; ;
Date Published:
Journal Name:
Membranes
Volume:
12
Issue:
2
ISSN:
2077-0375
Page Range / eLocation ID:
199
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ( , Water Environment Research)
    Abstract Practitioner Points

    This two‐decade UTEP‐EPW research partnership was sustained by a long‐term commitment to research and consistent financial support from EPW.

    Universities can collaborate to leverage utility funding toward larger external grant funding to advance research and development in a win–win partnership.

    The high‐recovery CERRO process was developed through multiple phases of concentrate management research, which would not have been possible without long‐term research commitment and risk tolerance from EPW.

    CERRO systems are being implemented at full scale in El Paso to recover water from silica‐saturated RO concentrate at an estimated specific energy consumption of 1.23 kWh/m3(4.6 kWh/kgal) and total amortized cost of $0.59/m3($2.25/kgal).

     
    more » « less
  2. ; ( , Environmental Science: Water Research & Technology)
    null (Ed.)
    Seawater desalination has become an important tool to attain global water security and sustainability. Among the available technologies, reverse osmosis (RO) has become the golden standard for seawater desalination due to its unparalleled energy efficiency. While RO is already efficient after development for half a century, there remains room for over 50% further reduction in energy consumption that can translate into tens of terawatt hours of potential annual energy saving. However, this significant energy saving cannot be achieved under the conventional paradigm of on-ground RO. In this analysis, we assess the idea of operating RO with open modules several hundred meters below the ocean surface ( i.e. , the mesopelagic zone). This new process, namely mesopelagic open reverse osmosis (MORO), can potentially push the energy consumption of seawater desalination to its theoretical limit. We first describe the concept of MORO, and then examine both the theoretical potential of energy saving and the practical challenges facing the implementation of MORO. Our analysis provides a theoretical framework for the future development of MORO for more sustainable desalination. 
    more » « less
  3. ; ; ; ; ; ; ; ; ; ( , Environmental Science: Water Research & Technology)
    Several areas around the world rely on seawater desalination to meet drinking water needs, but a detailed analysis of dissolved organic matter (DOM) changes and disinfection by-product (DBP) formation due to chlorination during the desalination processes has yet to be evaluated. To that end, DOM composition was analyzed in samples collected from a desalination plant using bulk measurements ( e.g. dissolved organic carbon, total dissolved nitrogen, total organic bromine), absorbance and fluorescence spectroscopy, and ultrahigh resolution mass spectrometry (HRMS). Water samples collected after chlorination ( e.g. post pretreatment (PT), reverse osmosis (RO) reject (brine wastewater) (BW), RO permeate (ROP), and drinking water (DW)), revealed that chlorination resulted in decreases in absorbance and increases in fluorescence apparent quantum yield spectra. All parameters measured were low or below detection in ROP and in DW. However, total solid phase extractable (Bond Elut Priority PolLutant (PPL) cartridges) organic bromine concentrations increased significantly in PT and BW samples and HRMS analysis revealed 392 molecular ions containing carbon, hydrogen, oxygen, bromine (CHOBr) and 107 molecular ions containing CHOBr + sulfur (CHOSBr) in BW PPL extracts. A network analysis between supposed DBP precursors suggested that the formation of CHOBr formulas could be explained largely by electrophilic substitution reactions, but also HOBr addition reactions. The reactions of sulfur containing compounds are more complex, and CHOSBr could possibly be due to the bromination of surfactant degradation products like sulfophenyl carboxylic acids (SPC) or even hydroxylated SPCs. Despite the identification of hundreds of DBPs, BW did not show any acute or chronic toxicity to mysid shrimp. High resolution MS/MS analysis was used to propose structures for highly abundant bromine-containing molecular formulas but given the complexity of DOM and DBPs found in this study, future work analyzing desalination samples during different times of year ( e.g. during algal blooms) and during different treatments is warranted. 
    more » « less
  4. ; ; ; ; ; ; ( , Water Resources Research)
    Abstract

    Applying models to developed agricultural regions remains a difficult problem because there are no existing modeling codes that represent both the complex physics of the hydrology and anthropogenic manipulations to water distribution and consumption. We apply an integrated groundwater – surface water and hydrologic river operations model to an irrigated river valley in northwestern Nevada/northern California, United States to evaluate the impacts of climate change on snow‐fed agricultural systems that use surface water and groundwater conjunctively. We explicitly represent individual surface water rights within the hydrologic model and allow the integrated code to change river diversions in response to earlier snowmelt runoff and water availability. Historically under‐used supplemental groundwater rights are dynamically activated within the model to offset diminished surface water deliveries. The model accounts for feedbacks between the natural hydrology and anthropogenic stresses, which is a first‐of‐its‐kind assessment of the impacts of climate change on individual water rights, and more broadly on river basin operations. Earlier snowmelt decreases annual surface water deliveries to all water rights, not just the junior water rights, owing to a lack of surface water storage in the upper river basin capable of capturing earlier runoff. Conversely, downstream irrigators with access to reservoir storage benefit from earlier runoff flowing past upstream points of diversion prior to the start of the irrigation season. Despite regional shifts toward greater reliance on groundwater for irrigation, crop consumption (a common surrogate for crop yield) decreases due to spatiotemporal changes in water supply that preferentially impact a subset of growers in the region.

     
    more » « less
  5. ; ; ; ; ( , Desalination)
    In-situ monitoring techniques of reverse osmosis (RO) desalination systems, particularly those with chemical sensing capabilities, can provide the means for better understanding important scaling mechanisms as well as early scaling detection. In this work, both calcium sulfate and calcium carbonate scaling on RO membranes were detected concurrently in real time using Raman spectroscopy to provide a unique chemical fingerprint. Two different sampling methodologies (manual and automated) were employed, and their performance was evaluated by comparing the Raman detection times to concurrent values of flux decline. The manual sampling strategy resulted in the detection of calcium sulfate and calcium carbonate at mean permeate flux declines of 13 ± 10 % and 22 ± 3 %, respectively. The automated sampling strategy provided better performance, with detection of calcium sulfate and calcium carbonate at mean flux declines of 8 ± 5 % and 4 ± 3 %, respectively. The increasedsensitivity and decreased variability of the automated sampling strategy provided valuable preliminary insights for the selection of optimized sampling strategies. The ability to identify the chemical composition of different scaling crystals including their polymorphs is an important step toward better understanding of the crystallization pathways of multi-component feed streams used in seawater and brackish water RO desalination. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email