A multiplexer pumping system to sequentially pump water from several sources to a unique automatic water quality probe
Participants: Marc Horstman, Beth Allen, François Birgand
Timeline: Started in 2009
Project Description
The main goal of this project is to be able to pump water samples
from multiple sources to a single sensor unit in the field. Because automatic continuous
water quality probes (e.g. s::can) are still very expensive, the idea is to expand the 'continuous' capabilities of these
probes to a spatial dimension by sequentially pumping and reading parameters such as nitrates, Dissolved Organic Carbon or turbidity from up to 10
different sources. We are essentially making a portable water quality lab that pumps waters of interest directly in-situ.
How does it work?
The multiplexor is driven by a programmable logic computer
(DL06 Micro PLC)
which send instructions, to for instance open a specific valve or to pump. All kinds of programmation can be done, on the picture you see an example of a
possible programmation. The multiplexer is equipped with ten valves, a peristatic pump (able to pump air and purge back) and a water sensor circuit,
to stop pumping when the water sample is ready to measure. Different kinds of projects have been tested to reduce cross contamination and dilution between
the ten ways of pumping. The Multiplexer is linked to 12V power battery for the PLC and probe alimentations.
The probe
The probe uses spectrometry principle to calculate concentrations
and turbidity. S::CAN spectro::lyzer measures
absorption at 256 different wavelengths. The wavelengths are dispersed across the range of 200 - 750 nm at intervals of approximately 2.5 nm for the UV-vis
probe. The Ana::pro Software displays and allows results editing on a computer. A fingerprint which represents the absorbance for known wavelenghts is
loaded: "Superposition of the absorption spectra of all substances present in the measured medium at the time of measuring." With this file and with a
previous calibration Ana::pro is able to determinate concentrations (NO3-N, DOC, TOC) and turbidity ,"using regression techniques that weight the
different wavelengths, according to what gives the best fit on a large set of data"
links
- Poster presentation of the intial prototype system
- Oral presentation of the system and of the early results
This page was last updated on April 1, 2011.
