What flow rates are supported for leak detection?

The flow rates suitable for leak detection depend on the technology used:

• Turbine Flow Sensors: These have a typical low-end limit of around 0.5 l/min due to factors related to fluid dynamics (see below). However, specialized algorithms like FlowReporter's "drip detection" (see below) can enhance accuracy down to  0.05 l/min, but this may increase detection time and potential for false alerts as the enhanced sensitivity for drips might cause false alerts from low-flow taps or drip irrigation systems. 

• High-Accuracy Water Meters (R500 or above):  These offer significantly better low-flow detection than turbine flow sensors. But, their resolution might be lower (pulses per volume), increasing leak detection times.

• Leak Detection Cable/Wireless Splash Sensors:  By far the most effective in detecting minuscule leaks, like drips, due to their direct contact with potential leak points.

Explaining Low-Flow Limitations of Turbine Flow Sensors

• Laminar Flow: At low flow rates, water follows a laminar flow pattern  The water travels faster in the pipe's centre due to lower friction. This non-linear velocity distribution reduces turbine flow sensor accuracy.

• Turbulent Flow:  At higher flow rates, turbulent flow creates tiny vortexes that even out  the water's velocity across the pipe. This improves turbine sensor accuracy. The transition point between laminar and turbulent flow is typically between 1-2 l/min for a 15mm pipe.

Optimising Detection

• Ideal Installation: To maximize turbine sensor accuracy at low flows, place sensors vertically with at least 10 pipe diameters of straight pipe upstream and 5 downstream. Minimize bends, tees, and other flow-disrupting components within this area.

Drip Detection

Go to Settings, you will then see a list of alerts, including a drip alert (<0.5 litres per minute):

Click on the alert to see what happened:

You will notice many small drip flows, almost unseen amongst the larger flow of water, but you can zoom in which makes them more noticeable:

The algorithm detects when there are more of these tiny spikes than usual in a particular time window. To enable these alerts, you must use Dynamic alerts and have clicked the Drip detection option when learning: