Identifying toxic materials in water with machine learning
Date:
March 21, 2022
Source:
University of British Columbia Okanagan campus
Summary:
Waste materials from oil sands extraction, stored in tailings ponds,
can pose a risk to the natural habitat and neighboring communities
when they leach into groundwater and surface ecosystems. Until now,
the challenge for the oil sands industry is that the proper analysis
of toxic waste materials has been difficult to achieve without
complex and lengthy testing. And there's a backlog. For example,
in Alberta alone, there are an estimated 1.4 billion cubic meters
of fluid tailings.
FULL STORY ========================================================================== Waste materials from oil sands extraction, stored in tailings ponds,
can pose a risk to the natural habitat and neighbouring communities
when they leach into groundwater and surface ecosystems. Until now,
the challenge for the oil sands industry is that the proper analysis of
toxic waste materials has been difficult to achieve without complex and
lengthy testing. And there's a backlog. For example, in Alberta alone,
there are an estimated 1.4 billion cubic metres of fluid tailings,
explains Nicola's Peleato, an Assistant Professor of Civil Engineering
at the University of British Columbia's Okanagan campus (UBCO).
==========================================================================
His team of researchers at UBCO's School of Engineering has uncovered a
new, faster and more reliable, method of analyzing these samples. It's
the first step, says Dr. Peleato, but the results look promising.
"Current methods require the use of expensive equipment and it can take
days or weeks to get results," he adds. "There is a need for a low-cost
method to monitor these waters more frequently as a way to protect public
and aquatic ecosystems." Along with masters student Mari'a Claudia
Rinco'n Remolina, the researchers used fluorescence spectroscopy to
quickly detect key toxins in the water. They also ran the results through
a modelling program that accurately predicts the composition of the water.
The composition can be used as a benchmark for further testing of other samples, Rinco'n explains. The researchers are using a convolutional
neural network that processes data in a grid-like topology, such as
an image. It's similar, she says, to the type of modelling used for
classifying hard to identify fingerprints, facial recognition and even self-driving cars.
"The modelling takes into account variability in the background of the
water quality and can separate hard to detect signals, and as a result
it can achieve highly accurate results," says Rinco'n.
The research looked at a mixture of organic compounds that are toxic,
including naphthenic acids -- which can be found in many petroleum
sources. By using high-dimensional fluorescence, the researchers can
identify most types of organic matter.
"The modelling method searches for key materials, and maps out the
sample's composition," explains Peleato. "The results of the initial
sample analysis are then processed through powerful image processing
models to accurately determine comprehensive results." While results to
date are encouraging, both Rinco'n and Dr. Peleato caution the technique
needs to be further evaluated at a larger scale -- at which point there
may be potential to incorporate screening of additional toxins.
Peleato explains this potential screening tool is the first step, but
it does have some limitations since not all toxins or naphthenic acids
can be detected -- only those that are fluorescent. And the technology
will have to be scaled up for future, more in-depth testing.
While it will not replace current analytical methods that are more
accurate, Dr. Peleato says this approach will allow the oil sands
industry to accurately screen and treat its waste materials. This is a necessary step to continue to meet the Canadian Council of Ministers of
the Environment standards and guidelines.
The research appears in the Journal of Hazardous Materials, and is
funded by the Natural Sciences and Engineering Research Council of Canada Discovery Grant program.
========================================================================== Story Source: Materials provided by University_of_British_Columbia_Okanagan_campus. Note: Content may be
edited for style and length.
========================================================================== Journal Reference:
1. Mari'a Claudia Rinco'n Remolina, Ziyu Li, Nicola's M. Peleato.
Application of machine learning methods for rapid fluorescence-based
detection of naphthenic acids and phenol in natural surface waters.
Journal of Hazardous Materials, 2022; 430: 128491 DOI: 10.1016/
j.jhazmat.2022.128491 ==========================================================================
Link to news story:
https://www.sciencedaily.com/releases/2022/03/220321091929.htm
--- up 3 weeks, 10 hours, 51 minutes
* Origin: -=> Castle Rock BBS <=- Now Husky HPT Powered! (1:317/3)