How might we detect and measure the taste and odour of water in real-time in a water treatment plant?

Solvents and organic compounds, such as 2-MIB, geosmin, acetone, and tetrahydrofuran, could exist in water infrequently at very low concentrations (i.e. parts per billion or even parts per trillion levels). Although they do not pose a health hazard and are within permissible drinking water quality standards, they can cause taste and odour issues at the consumer end and give rise to customer complaints. Furthermore, some of these compounds do not cause any taste and odour issues until they are boiled.

The industry’s current practice is to sample and measure these compounds in the laboratory using instruments, such as a purge-and-trap gas chromatography-mass spectrometer (P&T GC-MS) or a liquid chromatography-mass spectrometer (LC-MS) as part of a regular sampling and measurement regime. While real-time TOC sensors are also used to continuously monitor if these trace compounds are within stringent limits, they are not able to pinpoint the specific species of compounds from among the compounds present that may contribute to objectionable taste and odour.

We are interested in cost-effective, continuous-monitoring solutions that can detect and identify the specific odour and taste issues at the raw water intake to the plant, b​efore the water leaves the treatment plant, as well as in the network before it reaches the consumers.

We are interested in instruments or machines that can be implemented in a water treatment plant or service reservoir that allows rapid, specific, and real-time detection of taste and odour compounds. Some of the compounds may also be undetectable at ambient conditions but are only detected when they are heated or volatilised. We are seeking solutions that are not only able to discriminate between the various taste and odour-causing compounds but also to provide an indication of their concentration in the water in real-time.

• There are several known compounds such as 2-MIB, Geosmin, and other organic compounds, that can cause taste and odour issues even at very low levels such as in parts per billion levels or even parts per trillion.
• Despite their low levels, the human nose and taste buds can detect some of these compounds, solutions that are able to mimic these sensory systems or propose alternative ways to pick up taste and odour issues will be considered.
• The solution must be feasible to implement in a water treatment plant and at service reservoirs to monitor taste and odour (i.e. it should be able to operate in an automated manner).
• Not all the odorous compounds present in drinking water have been fully characterised, and their presence in drinking water is highly episodic with intervals of several years between episodes. Therefore, the solution should be capable of learning and adapting its detection ability to accommodate new target compounds.
• The cost of the solution should ideally be less than SGD 90,000 per unit.

After Selection

• Water from PUB water treatment plants for testing.
• Examples of compounds of concern and the concentrations at which they give rise to issues (For system validation).

Water Supply (Plants) Department & Water Quality Department

Total project period - less than 18 months
Development of “low fidelity” prototype - 9 months
Testing and validation of the solution in the field – 6-9 months

An integrated system that is able to pick up taste and odour issues, and sends an early alert to the operator so that remedial action can be taken.

Flexisolve Technology, Singapore