¹úÃñ²ÊƱ

Improving access to clean water in Fijian communities

Designing effective tools and models for water treatment authorities

Together with Water Authority of Fiji, ¹úÃñ²ÊƱ is working to improve water quality in Fiji by increasing the forecasting capabilities of the two water treatment plants.

¹úÃñ²ÊƱ design and modelling experts are working to establish a Hydrologic Information Management System (HIS) to enable the storage and maintenance of hydrological data, and a forecasting model to predict turbidity levels within the main tributaries to water treatment processing plants.

The status

The Water Authority of Fiji has two water treatment plants, situated within the Waimanu and Sigatoka catchments, which operate 24/7 at full capacity and at times are unable meet peak demand. Land use in the upper river catchments includes a mixture of forest, agriculture and mining, with changing land use and runoff causing turbidity within the river systems. The raw water is extracted from the Waimanu and Sigatoka rivers via offtake pumps, then entering the water treatment plants.

The challenge

At this stage, plant operators are not forewarned when turbid water is arriving, hence they have insufficient time to prepare. Service interruptions are more pronounced during periods of high rainfall, when the river waters become highly turbid, reducing the output of the plants as the filters require more frequent backwashing. The situation is compounded by lack of data on vast area of ungauged catchments and continuous threat of climate change, with expected changes in rainfall patterns and increased extreme rainfall events forecasting more frequent interruptions to water supply.

The goal

In collaboration with WAF, the ¹úÃñ²ÊƱ team is developing a remote sensing-land-based hydrological model that uses change of soil use and forecasted rainfall to determine the rate of turbidity in the river. Since there is lack of ground level measurements, the ¹úÃñ²ÊƱ team is incorporating smart remote sensing techniques in identifying change in river flow rate and in water quality when constructing the model framework. Finally, by using historical data from urban areas especially downstream of the catchments, we will be able to calibrate the models. Once calibrated, the models can act as a forecasting tool to WAF to predict incoming turbidity in the rivers by using forecasted rainfall events. The turbidity forecasting model will enable WAF to better manage chemical consumption, machine downtime and maintenance costs at the water treatment plants.

Project phases

  1. Phase 1: Project initiation – COMPLETE
  2. Phase 2: Establish Hydrologic Information System – COMPLETE
  3. Phase 3: Design conceptual models (including land based hydrological model) - April 2025
  4. Phase 4: Calibration of conceptual models - September 2025
  5. Phase 5: Model finalisation and reporting - December 2025
  6. Phase 6: Presentation and training - February 2026
  7. Phase 7: Incorporation of model results into WAF asset management and water security risk assessment  - April 2026