A gold-producing company with major operations in Africa also operates a gold tailings retreatment facility, recovering valuable residual minerals and turning waste into value. The plant receives its tailings from the company’s processing plants, located about 43 km / 26.7 mi away, via a large 20-inch transmission pipeline.

The tailings consist of a highly abrasive mixture containing chemicals, ore waste and large quantities of water, forming an aqueous slurry that can be pumped. Transporting the slurry, however, is difficult as its composition poses a significant risk to pipeline integrity. The high solids content and abrasive nature cause permanent wear at the pipeline, leading to recurring leaks throughout the year. Sabotage or theft, such as through the removal of bolts or other pipeline equipment, has also been a source of leaks.

When a leak occurs, the tailings slurry can “sandblast” the pipeline, increasing the size of the leak and rapidly causing a full pipeline rupture. Because the pipeline crosses areas of varying altitude, leaks can lead to uncontrolled slurry discharge, resulting in significant product loss, environmental damage and financial risk.

Leaks had so far only been detected once they were already big. As no effective leak detection and localisation was in place, the operator had resorted to time-consuming patrols along the entire pipeline route. This approach proved both ineffective and potentially dangerous to the customer’s personnel, as undetected leaks can rapidly turn the surrounding area into quicksand-like tailings ponds. In addition, patrolling in remote and unsafe locations, particularly at night, has its own occupational safety risks.

The mining company was therefore looking for a reliable pipeline leak detection and localisation solution to significantly improve reaction times and reduce the environmental impact of leaks. The pipeline transports approximately 13050 m³ of gold tailings per hour. Given that systems based on fibre optics, mass balancing or conventional real-time transient modelling had turned out to be ineffective and unsuitable, the operator had almost abandoned the search for a reliable and accurate leak detection system for mining slurries.

As a renowned instrumentation supplier to the mining industry for many decades, KROHNE was given the opportunity to test its PipePatrol NEO system on one of the tailings transfer pipelines. Having originally been developed for pipeline applications in the oil and gas industry, the leak detection and localisation system (LDS) has meanwhile been also adopted and optimised for slurries in the mining industry.

The system applies NEPM (Neural Engine Pipeline Monitoring), an AI-enhanced multi-method approach which integrates advanced hydraulic modelling, statistical pattern recognition and real-time measurements of flow, pressure and temperature. By continuously comparing a "virtual pipeline" model with actual operating data, the LDS detects deviations and analyses them using AI-based pattern recognition. It continuously learns from operational data, adapts to changing conditions and delivers precise, reliable pipeline information in real time.

With its real-time capabilities, PipePatrol NEO alerts operators within minutes of a pipeline leak, enabling an immediate response. It offers fast leak detection without the need for manual patrolling along the pipeline and helps mitigate the financial and operational risks associated with leaks through continuous monitoring, making PipePatrol NEO the solution of choice for safeguarding pipeline operations.

The system’s pattern recognition technology has been finely tuned to the leak behaviour of mining slurries. As leak signatures in slurry pipelines differ significantly from those of pipelines carrying other media, the LDS uses dedicated leak pattern recognition algorithms tailored to tailings and other mining slurries, making the system “slurry resistant”. This enables reliable, fast and accurate detection and localisation, even of small leaks, while ensuring extremely fast operator response times. Unlike many other leak detection systems, PipePatrol NEO is specifically designed to avoid false alarms, which can undermine the overall reliability of a leak detection system.

KROHNE supplied the full scope of the LDS, including software and hardware such as the LDS server and workstation. For continuous process monitoring and reporting via an intuitive HMI, the customer’s LDS workstation was equipped with SynEnergy, a supervisory and visualisation software.

For flow measurement OPTIFLUX 4300 electromagnetic flowmeters were installed. These flowmeters are available with rugged liners, e.g. soft rubber or ceramic tiles, both of which have a proven track record in mining applications involving extremely abrasive slurries with very high solid content. In addition, KROHNE supplied all necessary OPTIBAR PC 5060 pressure transmitters. These reliable and robust pressure transmitters feature a ceramic diaphragm, ideally suited for highly abrasive media.

The scope of supply entailed complete project management, including Site Acceptance Testing (SAT) and Performance Acceptance Testing (PAT), engineer training, and more. All components and instrumentation were seamlessly integrated into the IT infrastructure and SCADA system. The system utilises OPC UA communication standard for easy interoperability with the customer’s existing IT environment.

Accurate localisation drastically reduces clean-up costs. Gone are the days when patrolling along the entire pipeline route was necessary, with all the associated safety risks for the service personnel. The reduction in response time enables faster containment of product spillage and swift mitigation of environmental pollution, thereby minimising product loss and potential compensation payments.

In addition, collateral damage from leaks is effectively prevented, as clean-up and repair measures can be initiated promptly. The mining company is now able to detect leaks before sandblasting of pipelines occurs – an issue that would otherwise result in slurry flooding and the formation of quicksand-like tailings ponds within a short period. This helps protect adjacent equipment such as neighbouring pipelines or pumps, while reducing the risks for clean-up teams and equipment.

PipePatrol NEO has achieved what was long considered impossible: the accurate and reliable detection and localisation of leaks in tailings and slurry pipelines. Unlike other leak detection system suppliers, KROHNE has specifically tuned PipePatrol NEO to the unique characteristics of mining slurries. It provides a complete leak detection solution from a single source.

The combination of instrumentation expertise, data transmission and leak detection experience is unique in the industry, making KROHNE a full-scope supplier for demanding mining slurry leak detection. For customers requiring additional support, KROHNE also offers bespoke Service Level Agreements (SLAs) to assist with interpreting diagnostics and alarm signals.

PipePatrol NEO is modular and scalable, designed to grow with increasing operational requirements. Its hardware and software design allows for easy expansion and integration of additional pipelines. This flexibility has already paid off for the mining company, which – impressed by the system’s performance and precise alarms – has meanwhile extended the KROHNE LDS to two more tailings pipelines.