The challenge
High losses and no clear picture of the network
The water utility in Daruvar was losing a large share of its water and had no accurate, up-to-date picture of the distribution network. Without reliable data and a way to locate losses, it was difficult to plan effective interventions or justify investment.
The approach
Model, measurement and districting combined
Within an ECOLINKS-funded programme (USA), I combined desk modelling with intensive field work to find and reduce losses:
- Calibrated hydraulic model of the distribution network, built from digitised maps and elevation data
- Field measurements, hydrant tests and night-flow monitoring to detect and quantify leaks
- Network split into 6 measurement zones (and zone 4 further into sub-zones) to localise losses
- Integrated approach: hydraulic model + targeted flow/pressure measurement + classic leak-detection methods
~33 km
Total network length
1
Gravity source (inflow 19.8–49.4 l/s)
6
Measurement zones
30–65 m
Pressure range across the network
Leakage level · before vs after
Leakage by zone
| Zone | Length [km] | Leakage [l/s] | Leakage [m³/year] |
|---|---|---|---|
| 1 | 2.13 | 0 | 0 |
| 2 | 3.92 | 0.15 | 4 730 |
| 3 | 4.61 | 0.5 | 15 768 |
| 4a | 0.16 | 7.5 | 236 520 |
| 4b | 0.39 | 2.0 | 63 072 |
| 4c | 1.38 | 0 | 0 |
| 4d | 0.34 | 2.1 | 66 226 |
| 4e | 0.39 | 0.2 | 6 307 |
| 4f | 3.93 | 3.5 | 110 376 |
| 5 | 10.90 | 3.0 | 94 608 |
| 6 | 2.72 | 0 | 0 |
| Total | 30.85 | 18.95 | 597 607 |
Zone 4 (highlighted) concentrated the bulk of the losses — the basis for targeted, cost-effective leak repair.
The result
Losses brought down to about a third
Water losses were brought down to roughly a third of their original level, with 80% of the losses pinpointed on just 20% of the network length. The utility also gained a digital baseline of its network for day-to-day operation and future planning — turning a one-off project into a lasting operational asset.