6 Discussion
In the previous sections, a variety of spectroscopic and related optical methods has
been described, each with current applications in online water quality monitoring or
the potential to be used in such applications. These technologies all share the sole use
of interaction between light and matter as their principal measurement. The fully
solid-state sensor hardware and the lack of reagents mean these sensors are potentially highly robust and potentially provide long-term performance stability. The
current generation of sensors, however, is used in limited numbers and only very
rarely in larger numbers as would be expected in smart sensor networks. The main
reasons for their limited use are maintenance requirements, power requirements,
instrument price and approach to data interpretation.
All optical instruments have an interface where the light used for the analysis
crosses from the interior of the instrument into the sample and subsequently the same
interface or a secondary, for collection of the light and guiding it to the detector. As
only the interaction of the light with the sample is of interest, the optical interface
should be fully transparent. However, when it is in contact with the sample, there is a
risk of buildup of foreign material. Typical issues include scaling and (bio)fouling. It
is therefore critical that optical systems deal with these issues if they are to be
deployed in larger numbers, as otherwise the maintenance will be prohibitive. This
issue is currently not solved in a satisfactory manner. Although manual cleaning
intervals in drinking water applications are often satisfactory, in natural waters and
especially in wastewaters, these sensors require frequent (weekly–monthly) maintenance. New methods to prevent contamination of optical surfaces (e.g. antifouling
coatings) and/or methods to recognise and correct for fouling in the data processing
are required to deal with this issue.
Virtually all spectrometers use an artificial light source. The majority of the
advanced systems described in this chapter make use of incandescent lamps and
arc lamps. Examples are the xenon, deuterium, deuterium/halogen, tungsten/halogen
and mercury/argon lamps. Except for the xenon lamp, which is often used as a flash
lamp, these lamps are used in continuous mode. As the lifetime of these lamps is
measured in hundreds or thousands of hours, they need replacement. Furthermore,
the power requirements for these lamps are such that a main power supply is required
to operate the instruments. LED technology provides an alternative to these lamps.
Although not all relevant wavelengths are currently available using LEDs, UV
spectroscopy and fluorescence devices are making use of this technology. Although
LEDs offer an advantage regarding cost and power consumption, allowing for longterm battery-powered operation, they suffer from decreasing brightness over time,
which needs to be monitored and corrected for. As a result such instruments require
occasional recalibration or replacements of LEDs, e.g. every 1–2 years.
Instrument prices for the devices described vary widely but are all in the 1,000+
euro range. The simplest LED-powered single or dual wavelength devices can be
acquired for a few thousand euro, whereas the more advanced spectrometers may
cost upwards of 30,000 euro. Even the costs for the simpler instruments prohibit
308
J. van den Broeke and T. Koster
In the previous sections, a variety of spectroscopic and related optical methods has
been described, each with current applications in online water quality monitoring or
the potential to be used in such applications. These technologies all share the sole use
of interaction between light and matter as their principal measurement. The fully
solid-state sensor hardware and the lack of reagents mean these sensors are potentially highly robust and potentially provide long-term performance stability. The
current generation of sensors, however, is used in limited numbers and only very
rarely in larger numbers as would be expected in smart sensor networks. The main
reasons for their limited use are maintenance requirements, power requirements,
instrument price and approach to data interpretation.
All optical instruments have an interface where the light used for the analysis
crosses from the interior of the instrument into the sample and subsequently the same
interface or a secondary, for collection of the light and guiding it to the detector. As
only the interaction of the light with the sample is of interest, the optical interface
should be fully transparent. However, when it is in contact with the sample, there is a
risk of buildup of foreign material. Typical issues include scaling and (bio)fouling. It
is therefore critical that optical systems deal with these issues if they are to be
deployed in larger numbers, as otherwise the maintenance will be prohibitive. This
issue is currently not solved in a satisfactory manner. Although manual cleaning
intervals in drinking water applications are often satisfactory, in natural waters and
especially in wastewaters, these sensors require frequent (weekly–monthly) maintenance. New methods to prevent contamination of optical surfaces (e.g. antifouling
coatings) and/or methods to recognise and correct for fouling in the data processing
are required to deal with this issue.
Virtually all spectrometers use an artificial light source. The majority of the
advanced systems described in this chapter make use of incandescent lamps and
arc lamps. Examples are the xenon, deuterium, deuterium/halogen, tungsten/halogen
and mercury/argon lamps. Except for the xenon lamp, which is often used as a flash
lamp, these lamps are used in continuous mode. As the lifetime of these lamps is
measured in hundreds or thousands of hours, they need replacement. Furthermore,
the power requirements for these lamps are such that a main power supply is required
to operate the instruments. LED technology provides an alternative to these lamps.
Although not all relevant wavelengths are currently available using LEDs, UV
spectroscopy and fluorescence devices are making use of this technology. Although
LEDs offer an advantage regarding cost and power consumption, allowing for longterm battery-powered operation, they suffer from decreasing brightness over time,
which needs to be monitored and corrected for. As a result such instruments require
occasional recalibration or replacements of LEDs, e.g. every 1–2 years.
Instrument prices for the devices described vary widely but are all in the 1,000+
euro range. The simplest LED-powered single or dual wavelength devices can be
acquired for a few thousand euro, whereas the more advanced spectrometers may
cost upwards of 30,000 euro. Even the costs for the simpler instruments prohibit
308
J. van den Broeke and T. Koster