Designing Suspended-Solids Feedback for Dredge Control

Automatic dredging systems increasingly use optical measurements to control production rate, pump output, cutter speed, overflow behaviour, or vessel movement. The aim is to keep sediment plumes within an approved operating envelope while maintaining useful excavation rates. A suspended-solids sensor can provide the real-time signal needed for that balance, but only when its measurement, placement, communications, and control logic are designed as one system.

Integrating suspended-solids sensors into dredge automatic feedback control loops requires more than connecting an instrument to a programmable logic controller. Dredged material changes rapidly, water chemistry affects optical readings, and the same concentration can produce different signals when particle size or mineral composition changes. A robust installation therefore combines field calibration, hydraulic understanding, signal conditioning, and conservative control decisions suited to Australian marine and freshwater conditions.

Define the process variable before selecting the sensor

The first decision is what the control loop must regulate. Turbidity is an optical response, usually expressed in nephelometric turbidity units, while suspended solids concentration describes the mass of particles in a volume of water, commonly in milligrams per litre. These variables are related at a particular site, but they are not interchangeable under every sediment condition. A clay-rich plume may produce a very different optical response from coarse quartz sand at the same mass concentration.

If the environmental approval specifies turbidity limits, a turbidity monitor may be the most direct feedback instrument. If the dredge operator needs a mass-based estimate for production accounting or discharge management, the sensor output should be related to laboratory suspended-solids results through a site-specific correlation. The relationship should be established using samples collected during representative dredging conditions, including different stages of the tide and different material layers.

Unit conversion can also create avoidable confusion in a control system. Nephelometric values depend on the optical method, calibration material, instrument geometry, and reporting convention. Before mapping a sensor value to a PLC alarm or a compliance threshold, review turbidity conversion guidance and document the units used by every device in the chain.

Place the measurement where it represents the plume

Sensor location determines whether the feedback signal describes the dredging process or merely local turbulence. A monitor mounted too close to a cutter head can experience bubbles, rapidly changing solids loading, and direct impacts from large particles. It may react quickly, but its output can be too erratic for stable automatic control. A sensor positioned too far downstream may provide a smoother signal while adding transport delay that makes the loop slow or unstable.

The preferred location depends on the dredge type and the control objective. A trailing suction hopper dredger may require monitoring near overflow discharge, at a controlled outfall, or along the plume boundary. A cutter suction dredger may use a sensor near the discharge pipeline, at a return-water location, or on a monitoring frame outside the immediate excavation zone. Multiple sensors can distinguish process concentration from receiving-environment concentration, although each additional measurement point adds maintenance and data-validation requirements.

Australian sites make positioning particularly important. Work near the Port of Brisbane can involve strong tidal exchange and suspended sediment already present in the navigation channel. In Darwin, wet-season runoff and large tidal ranges may alter background conditions quickly. A sensor frame designed for one site should not be transferred to another without checking depth, current velocity, biofouling exposure, access arrangements, and the likely particle-size distribution.

Build a control loop around real sensor behaviour

A useful dredge control loop has four functional stages: measurement, validation, decision, and action. The optical sensor produces a reading, the control system checks whether the reading is plausible, the control algorithm compares it with a target or alarm band, and the dredge changes an operating parameter. Depending on the equipment, the manipulated variable may be cutter speed, swing speed, pump speed, production rate, overflow duration, or the position of a discharge valve.

Signal filtering should remove short-lived spikes without hiding a genuine plume increase. A moving average or first-order filter is often suitable, but its time constant must be compared with the travel time between the process and the sensor. If water takes 30 seconds to reach the measurement point and the controller waits another 60 seconds before acting, an aggressive response can create oscillation. A rate limiter and a deadband can prevent the machinery from repeatedly increasing and decreasing output around a threshold.

Control actions should also be graduated. A first response might reduce excavation rate slightly, followed by a stronger reduction if the concentration remains high. An independent high-high alarm can pause the relevant operation or notify the dredge master. This structure is safer than allowing a single noisy reading to command an abrupt stop, while still preserving a defined response when the plume is persistent.

Manage calibration and changing sediment conditions

Optical suspended-solids sensors are valuable because they provide frequent measurements without continuous laboratory handling, but the optical signal is influenced by particle size, shape, colour, refractive index, and concentration range. Calibration should therefore use samples from the actual dredging area. A single grab sample taken during mobilisation is unlikely to describe the full range encountered during a campaign.

A practical calibration programme pairs sensor readings with laboratory results over low, medium, and high concentration conditions. Samples should be collected at the same point and as close as possible to the same time as the optical measurement. The resulting regression, lookup table, or segmented calibration can then be loaded into the supervisory system. When sediment layers change, the project team should be able to flag the new material and apply a revised relationship without losing the original data record.

Routine checks are essential in Australian waters, where warm conditions can encourage marine growth and long campaigns may expose instruments to fine sediment abrasion. Wipers, compressed-air cleaning, copper components, mechanical guards, or scheduled manual cleaning can reduce fouling, but no protection is maintenance-free. The control system should identify a fouled or unstable sensor rather than treating its output as a real environmental increase.

Connect the instrument to automation infrastructure

The integration layer must preserve the quality and meaning of the measurement. Depending on the sensor and site architecture, communication may use analogue current, serial communications, Ethernet, or an industrial protocol. The PLC or remote terminal unit should receive the measured value, units, timestamp, diagnostic state, quality flag, and—where available—an indication of fouling, optical saturation, or communication failure.

A raw value without status information is unsuitable for automatic control. The PLC should distinguish between a valid high concentration and a frozen value caused by a broken cable. It should also detect out-of-range readings, implausible rates of change, loss of power, and stale communications. If the measurement becomes invalid, the loop should move to a defined fallback mode, such as holding the last safe production setting, reducing the dredging rate, or transferring control to the operator.

Integration should include a data historian or SCADA display that shows the sensor value alongside dredge position, pump status, production rate, tide, rainfall, and operator actions. This record helps explain why a control action occurred and supports environmental reporting. It is particularly useful for projects near the Great Barrier Reef, where marine-water quality monitoring may need to be interpreted alongside weather, river discharge, and vessel activity rather than viewed as an isolated number.

D & A Instruments equipment is now supported by Campbell Scientific, so system designers can obtain current product and integration details through the product information team. Confirming interface options, environmental ratings, mounting arrangements, and compatible accessories before procurement can prevent late changes to the PLC panel or vessel network.

Verify performance during commissioning

Commissioning should begin with observation rather than automatic intervention. Run the sensor and control display while the dredge operates under normal conditions, then compare the live signal with water samples, visual plume observations, and nearby monitoring stations where available. This period reveals delays, baseline variation, cross-sensitivity to bubbles, and differences between the sensor location and the compliance location.

Once the measurement is credible, test the feedback loop in stages. Simulate high and low readings, disconnect the communications path, introduce a sensor-quality fault, and check that the machinery responds within the intended limits. Verify that an operator can override automatic control, that the override is logged, and that the system returns to automatic mode only through a deliberate action. Alarm priorities should be clear enough for a dredge master working in a noisy control room, especially during night operations or poor weather.

Acceptance testing should use realistic response times. A loop that behaves well with a simulated instant signal may behave differently when the plume takes several minutes to move from the dredge to the monitoring frame. Record the delay, filter setting, control limits, and resulting production response in the commissioning documentation. This information becomes the baseline for later tuning and for explaining any changes to the project owner or environmental authority.

Keep operators and environmental teams in the loop

Automatic control works best when it supports operational judgement rather than hiding the reason for a decision. The operator should see the current concentration or turbidity, the configured limit, the sensor-quality state, the active control mode, and the action being applied. A simple display can show whether the system is responding to a genuine trend, a temporary spike, or a failed measurement.

Operating procedures should define what happens when weather changes, background turbidity rises, or the dredge moves between material types. In freshwater work around the Murray–Darling Basin, rainfall-driven inflows can alter background suspended solids independently of dredging. In urban harbour projects around Sydney, vessel movements and propeller wash may create short disturbances that should be recorded but not automatically treated as a dredging failure. These conditions call for documented hold points and review rules rather than arbitrary alarm suppression.

Maintenance and troubleshooting information should be accessible to both the electrical team and the environmental monitoring team. The technical support resources can help establish the relevant instrument documentation and support pathway, while site personnel should retain records of cleaning, calibration checks, sample results, and firmware or configuration changes. A sensor becomes a dependable control input when its history is understood, not simply when it produces a number.

With suitable placement, verified calibration, fault-aware communications, and measured control actions, suspended-solids monitoring can reduce plume risk without unnecessarily sacrificing dredging productivity. The result is a feedback system that reflects the realities of Australian waterways and gives operators a defensible, traceable basis for automatic decisions.