TOWARDS FAST RESPONSE TO HARMFUL BACTERIA LEVELS IN N EARSHORE WATERS: A LOCAL HYDRODYNAMIC MODEL AND IN SITU BIOSENSOR
Progress Report
Period of Performance: April 1st 2017 – March 31st 2018
Contact information:
Principal Investigator: Sylvia Rodríguez-Abudo, Ph.D. Department of Engineering Sciences and Materials Email: rodriguez.abudo @upr.edu
Co-Principal Investigator: Pedro Resto Irizarry, Ph.D. Department of Mechanical Engineering Email: pedroj.resto@upr.edu
I. Results and findings
Hydrodynamics and Water Quality Modeling:
We have successfully implemented wave forcing into the model using Delft3D SWAN. The model is now capable of simulating nearshore hydrodynamics in response to tidal, wind, and wave forcing. Comparisons between Delft3D SWAN and CARICOOS SWAN yield good agreement in both, wave height and direction (Figure 1). The root-mean-squared difference (RMSD) is consistently lower in the beach, which is encouraging given that this is the area where the beach water quality modeling will eventually take place. Additionally the spatial distribution of wave heights (Figure 2) suggests that the model is capable of reproducing wave transformation reasonably, as seen in the blue shadow located at the reef canyon.
Figure 2: Spatial distribution of wave heights (colorbar) as resolved by Delft3D
Issues to report: The results presented in this section date back to May 2017. No additional progress has been achieved ever since, as graduate student Daniel Martínez stopped working for the project starting Summer 2017, when he accepted an internship at USACE’s ERDC, followed by a position at the same national laboratory (October 2017). To mitigate this situation the PI has asked for permission to the program manager and export control to initiate a collaboration with TUDelft and Deltares, the developers of the Delft3D model, to continue with the project as soon as possible. On April 5th, 2018 we got the authorization from Dr. Yulissa García (Executive Officer, Sea Grant College Program) and Héctor Segarra (UPRM Export Control Officer) to initiate the collaboration.
In Situ Biosensor:
Students had previously constructed a portable fecal bacteria biosensor. The biosensor had been tested using freshwater samples inoculated with enterococcus faecalis pellets of a known CFU. The biosensor had been used for experiments on an almost weekly basis. Unfortunately, the stress caused on the Raspberry Pi mini-computer by the relatively large amounts of current used for controlling the CMOS camera and the hot plate meant that all of the CMOS cameras at our disposal broke down. During the past months we moved the hot plate control transistors off-chip to an electronics board outside of the Raspberry Pi. We had to purchase new CMOS sensors to continue our work of obtaining automated fluorescence images. We also built a new acrylic millifluidic chip and tested it for flow. M.S. student Katerin Rodriguez will be in charge of adapting a new electronic setup to control the 12V peristaltic pump and valves that will be used to control liquid loading and unloading in the millifluidic device. We are in the process of recreating the 24-hour incubation experiment. Future work will include performing automated water loading and unloading from the millifluidic device. We collaborated with Marine Sciences graduate student, Geraldine Gomez, to use our biosensor with seawater samples obtained from various beach locations. We used excess seawater from Geraldine’s experiments to compare our chip to the Quantitray. The goal was to obtain qualitative data comparing our chip to the Quantitray and study the feasibility of our chip to obtain quantitative data. II. Methods
The methodology for the hydrodynamic simulations remains the same. Delft3D simulations are carried out with a time step of 1 second, state-of-the-art curvilinear grid with coastal resolution of 15-20m, and 16 varying vertical sigma layers. The hydrodynamic grid is designed to cover great extent of the considered offshore zone (250-300 m depth) with cell resolution of 50-60 m. Tidal forcing (water levels) were input into the model using MATLAB’s Tidal Model Driver. Meteorological information obtained from the CARICOOS San Juan buoy, located approximately 20 km from Dorado, were used to implement the wind forcing. The model setup contains a spin-up time of 72 hours, where the hydrodynamic boundary conditions and the wind input signal are treated in manner to evolve smoothly by applying a hyperbolic weighting function.
Similarly, the methodology for the development of the in situ biosensor remains unchanged. The following describes the methods, instruments and devices used for the initial development of the in situ biosensor:
• Open source programming using Raspberry Pi micro-computer and Arduino microcontroller.
• Open source and low-cost rapid prototyping using a laser cutter and a Lulzbot 3D printer.
• Electronics work and Python programming designed and constructed by students.
• Heat control using a 3D printer hot plate and a One-Wire programmable digital thermometer. System calibration includes temperature versus time plots within the millifluidic device, at the location of liquid, and at the hot plate surface.
• Fluorescence analysis using Python programming and CMOS camera. The CMOS camera is being used with the acylic millifluidic device loaded with an Enterolert assay inoculated with a known Enterococcus faecalis CFU
III. Objectives
As stated in the original proposal, the objectives of the present project are as follows:
1. Numerical simulation of site-specific nearshore hydrodynamics and transport, concentration, and residence time of fecal indicator bacteria: We have implemented wave forcing into the model, which is now capable of simulating nearshore hydrodynamics in response to wind, waves and tides. With the departure of graduate student Daniel Martínez, the hydrodynamic simulations have been on hold. We have reach out to collaborators in TUDelft and Deltares, who will aid in the continuation of the project. We have been granted permission by Sea Grant and Export Controls to initiate the process.
2. Development of first generation in-situ biosensor for quasi-real time detection of fecal indicator bacteria: We previously constructed a portable bacteria biosensor with 24-hour temperature incubation and automated acquisition of fluorescence measurements using a CMOS camera. We have done 24-hour incubations of our device loaded with inoculated water and the Enterolert assay reagents in an incubator, and have measured positive and negative Enterolert assay results with the CMOS camera. We will continue to test our system by performing more 24-hour incubations of Enterolert assays within the
millifluidic device using inoculated water samples. A new electronic setup will be built to control 12V peristaltic pump and valves for automated liquid loading and unloading. The biosensor will be tested with real-world seawater samples and compared to Quantitray measurements.
IV. Other products
o No additional products yet.
V. Students and personnel supported
Students supported during the reporting period:
Daniel Martínez, daniel.martinez4@upr.edu, MS student Mechanical Engineering Sea Grant Research Assistantship paid Mr. Martínez $1293.75 for the period he worked for the project. During this time (Spring semester 2017) Mr. Martínez also received a Teaching Assistantship by the Department of Engineering Sciences and Materials.
Dong Wang, dong.wang@upr.edu, MS student Mechanical Engineering Sea Grant Research
Category # of new students # of continuing students # of degrees awarded
Grant
Grant
PhD
Grant
Assistantship has paid Mr. Wang $15,284.65 so far. He received a Teaching Assistantship by the Department of Mechanical Engineering until the Fall 2017-18 academic semester.
Thesis and dissertations by supported students
• Dong Wang, MS Thesis, ongoing: An Automated In-Situ Biosensor for Quasi Real-time Detection of Fecal Bacteria in Recreational Waters (tentative title)
List presentations, technical reports and special awards. (Send copies when available)
Martínez, D. and S. Rodriguez, Towards Beach Water Quality Modelling in Dorado PR: Preliminary Hydrodynamic Simulations Poster presented at the 2017 CARICOOS General Assembly in Añasco, PR.
Wang, D. and P. Resto, Automated, Portable Fecal Bacteria Biosensor for Recreational Water. Poster presented at the 2017 CARICOOS General Assembly in Añasco, PR.
List references for, books, chapters, and peer reviewed publications, in press, and submittals.
N/A
Time & Effort
The following tables list the effort level of the researchers, graduate and undergraduate students involved in the present project from April 1, 2017 – March 31, 2018
Researchers:
Sylvia Rodriguez
Graduate students:
VI. Impact/Accomplishment Statement
With the goal of increasing our understanding of the fate and transport of fecal indicator bacteria (FIB) in beach water, a local hydrodynamic model and bench-scale in situ biosensor are currently under development. The model will form the basis for water quality simulations under different oceanographic and meteorological scenarios, while the sensor will provide for increased frequency of FIB sampling and decreased incubation times. Currently the model is able to simulate nearshore hydrodynamics in response to winds, waves, and tides. The biosensor has been successfully constructed and is now under continuous testing. This project responds to increased exceedances of FIB contamination in the local area of Dorado Public Beach, PR. It brings together expertise in nearshore processes and microfluidics to ultimately provide local authorities with tools for effective management of public beaches.