Online water quality monitoring: why TDS alone is not enough? — Engineering practice for industrial multi-parameter monitoring

Introduction

The article "The lower the TDS value, the safer the water? 99% of people have misunderstood" profoundly reveals the common misunderstanding of TDS (Total Dissolved Solids) among the public. However, in professional fields such as industrial process control, municipal water supply and drainage, and environmental monitoring, if this misunderstanding is carried into the selection and design of online monitoring systems, it can lead to serious monitoring blind spots, missed alarm risks, and even process failures or public health incidents. TDS is only one dimension of water quality parameters. Online water quality monitoring needs to build a multi-parameter collaborative sensing system to comprehensively and timely reflect water quality changes, ensuring water supply safety and process stability.

This article is intended for water treatment engineers, system integrators, and managers responsible for water quality monitoring. From an engineering implementation perspective, it systematically explains why online monitoring cannot focus only on TDS, which key parameters need attention, how to select and integrate multi-parameter sensors based on application scenarios, and provides practical guidance for validation and maintenance.

1. Current Status and Limitations of TDS Online Monitoring

1.1 Engineering Definition and Online Measurement of TDS

TDS (Total Dissolved Solids) refers to the total amount of dissolved solid substances in water, mainly including inorganic salts (carbonates, chlorides, and sulfates of calcium, magnesium, sodium, and potassium) and a small amount of organic matter. Online TDS monitoring typically uses the conductivity method: measuring the conductivity of a water sample through an immersion or flow-through conductivity electrode, and then converting it to a TDS value based on an empirical coefficient (e.g., 0.5–0.7). This method offers fast response, simple maintenance, and low cost, making it widely used in scenarios sensitive to salt content, such as industrial cooling water, reverse osmosis (RO) product water, and boiler feedwater.

1.2 Water Quality Risks Not Reflected by TDS

However, the "safety" illusion of TDS is precisely the biggest engineering trap. Relying solely on the TDS indicator will miss the following important water quality risks:

• Microbial contamination: TDS does not reflect the presence of pathogenic microorganisms such as bacteria, viruses, and algae. Even water with very low TDS may face serious microbial risks due to secondary contamination in the distribution network or raw water emergencies.
• Organic pollutants: Natural organic matter (e.g., humic acid), industrial organics (e.g., phenol, pesticides), and disinfection by-products (e.g., trihalomethanes) have direct or indirect harmful effects on human health, but their contribution to conductivity is usually small and cannot be captured by TDS.
• Particulate matter and turbidity: Suspended particles (sediment, rust) and colloids affect not only sensory perception but also wrap microorganisms and weaken disinfection effects. However, they are non-conductive, so TDS values show no response.
• Toxic heavy metals: Some heavy metal ions (e.g., lead, mercury, cadmium) can be toxic at low concentrations. Although they increase conductivity, TDS is a total quantity indicator and cannot distinguish non-toxic ions (e.g., calcium, magnesium) from toxic ions.
• Disinfectant residual: Residual chlorine and total chlorine protect the pipeline system, but in certain scenarios (e.g., RO membranes), strict control is required, and TDS cannot indicate this.

1.3 A Typical Engineering Lesson

A beverage bottling plant used reverse osmosis product water with online monitoring of only TDS (expected value < 10 mg/L). One day, due to organic contamination in the raw water, excessive coagulant was dosed before the RO membrane, resulting in a decrease in product water pH and an increase in aluminum ions. However, the TDS value remained within the normal range because other ions were removed. As a result, flocculent precipitates appeared in the product, leading to batch rejection. Post-analysis showed that if turbidity, pH, and aluminum (or alternative parameters such as conductivity combined with pH) had been monitored online, early warning could have been provided. This clearly demonstrates that TDS is only "necessary but not sufficient" in comprehensive water quality assessment.

2. "Other Eyes" That Online Water Quality Monitoring Must Consider

A complete online water quality monitoring system needs to scientifically select parameter combinations based on different application goals. The following are core online monitoring parameters that complement TDS, and their irreplaceable reasons.

2.1 Turbidity

• Engineering significance: Turbidity is a measure of the amount of suspended particulate matter in water and also serves as an "indicator" of microbial risk. Particles can shelter pathogens, interfere with disinfection, and may carry adsorbed heavy metals or organic matter.
• Online monitoring method: 90° scattered light method (ISO 7027) or transmitted light method. Modern online turbidity meters can output NTU values in real time, responding to particle fluctuations.
• Complementarity with TDS: TDS reflects the amount of dissolved solids, while turbidity reflects the amount of suspended particles. Together, they form a complete picture of "solids" in water.

2.2 Comprehensive Organic Matter Indicators (COD, TOC, UV254)

• COD (Chemical Oxygen Demand): Represents the amount of organic matter (including some inorganic substances) in water that can be oxidized by strong oxidants. It is a classic indicator of organic pollution, but laboratory testing is time-consuming.
• TOC (Total Organic Carbon): Directly measures organic carbon content, which is more fundamental than COD, but online equipment costs are higher.
• UV254: Utilizes UV absorbance at 254 nm, highly correlated with aromatic organic matter (e.g., humic substances) and precursors of potential disinfection by-products. Online measurement is simple and fast.
• Online monitoring implementation: Multi-spectral probes (such as Oromë NSDD6) use non-contact full-spectrum scanning and algorithms to simultaneously output organic parameters such as COD, TOC, and UV254, without chemical reagents, avoiding secondary pollution and the maintenance hassle of frequent reagent replacement.
• Complementarity with TDS: Organic matter hardly affects conductivity. Therefore, even when TDS is stable, if the water source is invaded by organic matter, COD/TOC will rise sharply, providing much more sensitive early warning capability than TDS.

2.3 Color

• Definition: True color is the color of water after removing suspended solids, mainly derived from dissolved organic matter (such as humic acid) and industrial dyes.
• Online measurement: Absorbance in the visible light band, one of the standard outputs of multi-spectral sensors.
• Engineering value: Sensory acceptance is a key quality indicator in the drinking water and food industries. A sudden increase in color often indicates process abnormalities or raw water changes, and colored organic matter may react with disinfectants to form harmful substances.

2.4 Conductivity (EC) and Reunderstanding TDS

• EC is superior to TDS in online monitoring: Conductivity is the raw parameter without conversion, avoiding the empirical error of TDS conversion coefficients. Therefore, in industrial control, conductivity (μS/cm) is often directly used as a setpoint.
• Derived parameters: Temperature-compensated conductivity can be used to calculate salinity, which is used in seawater desalination and aquaculture; combined with density models, it can also estimate specific gravity for chemical processes.
• Integration example: The Oromë 5-in-1 conductivity sensor outputs EC, TDS, salinity, specific gravity, and temperature from a single probe. The pure titanium electrode is corrosion-resistant and meets continuous online requirements, avoiding the mechanical complexity and mutual interference of multiple probe installations.

2.5 Temperature

Although temperature is often regarded as a "auxiliary parameter," its role should not be underestimated:

• Effect on conductivity: Conductivity changes by approximately 2% for every 1°C change in temperature, so all conductivity/TDS sensors require temperature compensation.
• Effect on chemical reactions: The formation of disinfection by-products, coagulation and sedimentation efficiency, and membrane flux are all closely related to water temperature.
• Ecological impact: In aquaculture and surface water monitoring, temperature is a fundamental indicator of ecosystem health.

2.6 Other Important Parameters (Depending on Scenario)

• pH: Affects metal corrosion, disinfectant efficacy, and biological treatment efficiency.
• Residual chlorine/total chlorine: Ensures the disinfection capability of the pipeline network.
• Dissolved oxygen: Indicates water self-purification capacity and is used in aquaculture.
• Ammonia nitrogen and nitrate: Related to eutrophication and drinking water safety.

Among the above parameters, turbidity, organic matter, color, conductivity/salinity, and TDS together form the "multi-parameter cornerstone," covering the real-time monitoring needs of most industrial and municipal water treatment scenarios.

3. Engineering Necessity of Multi-Parameter Online Monitoring

3.1 From Single-Parameter Instruments to Multi-Parameter Systems

Early water treatment control often involved independently installing multiple single-parameter instruments: turbidity meters, conductivity meters, pH meters, thermometers... This led to messy cables, many installation joints, multiple maintenance points, and data silos. Modern integrated multi-parameter sensors (such as NSDD6, which achieves six organic/optical indicators with one probe, and the 5-in-1 probe, which achieves five electrochemical indicators) greatly simplify the system architecture:

• Reduced installation points: A single probe with multiple indicators reduces construction and civil engineering costs.
• Data from the same source: All measurements are taken under the same flow conditions, providing stronger correlations and facilitating intelligent early warning.
• Centralized maintenance: Only one or very few probes need maintenance, reducing labor.

3.2 Correlated Alarming and Traceability

Multi-parameter combinations can achieve intelligent diagnosis that single parameters cannot. For example:

• Raw water pollution traceability: If TDS is normal but turbidity and UV254 rise sharply, it is likely organic particulate pollution (e.g., humus after storm runoff); if only turbidity increases while UV254 remains stable, it may be inorganic particles (e.g., sediment).
• Membrane process abnormality: A slow increase in RO product water conductivity usually indicates membrane damage or seal leakage. If accompanied by increased product water turbidity and TOC, it confirms membrane integrity failure, requiring immediate shutdown for maintenance.
• Distribution network water quality model: Combining online turbidity, residual chlorine, pH, and TDS can infer the risk of pipeline corrosion and microbial regrowth.

3.3 Regulatory and Standard Drivers

Internationally, the World Health Organization (WHO) Guidelines for Drinking-water Quality and national drinking water standards (e.g., China's GB 5749-2022) set limits on multiple indicators. Although online monitoring data cannot be directly used as enforcement evidence, continuous monitoring helps water plants adjust processes before exceeding limits, avoiding compliance risks. Many smart water projects have already configured multi-parameter online micro-stations as standard equipment.

4. Multi-Parameter Sensor Selection and Integration Practice

4.1 Application Scenarios and Parameter Requirement Matrix

The first step in selection is to clarify the monitoring scenario and objectives. The table below sorts out common scenarios and recommended online parameter combinations ("√" indicates strongly recommended).

| Application Scenario | Turbidity | Organic Matter (UV254/COD/TOC) | Color | Conductivity/TDS/Salinity | Temperature | Other Key Parameters | |----------------------|------|----------------------|------|----------------|------|----------------| | Municipal drinking water treatment (inlet/outlet) | √ | √ | √ | √ | √ | pH, residual chlorine, turbidity | | Community secondary water supply | √ | √ (optional) | √ | √ | √ | Residual chlorine | | Reverse osmosis (RO) system | √ | √ (TOC) | | √ | √ | pH, ORP | | Industrial wastewater treatment | √ | √ (COD/TOC) | √ | √ | √ | pH, ammonia nitrogen | | Surface water/environmental monitoring | √ | √ (TOC/UV254) | √ | √ | √ | Dissolved oxygen, ammonia nitrogen | | Aquaculture | √ | | √ | √ (salinity) | √ | Dissolved oxygen, pH | | Boiler feedwater | | √ (TOC) | | √ | √ | Dissolved oxygen, pH |

4.2 Key Points for Sensor Technology Selection

After selecting parameters, sensor types need to be chosen based on installation environment, maintenance conditions, and technical requirements.

#### 4.2.1 Optical vs. Electrochemical Methods

• Organic matter and turbidity: Multi-spectral optical method is a trend, offering no reagent requirement, fast response, and low maintenance. Focus on the following:
• Light source: LED or xenon lamp life and stability;
• Self-cleaning capability: In sewage or high-turbidity water, automatic physical wipers or ultrasonic cleaning can effectively reduce manual maintenance frequency (e.g., NSDD6 is equipped with an automatic physical cleaning device);
• Material: 316L stainless steel, POM, and other corrosion-resistant materials suitable for immersion installation.
• Conductivity/TDS: When selecting electrode-type sensors, pay attention to the following:
• Electrode material: Pure titanium or graphite electrodes are corrosion-resistant with long life (5-in-1 uses pure titanium electrodes);
• Temperature compensation method: Built-in temperature sensor for automatic compensation;
• Flow cell design: Industrial processes may require flow-through installation to reduce maintenance.

#### 4.2.2 Communication and Integration

• Digital interface: Prefer RS-485 Modbus RTU protocol for easy integration into PLC or SCADA systems. Isolated communication over long distances is a basic requirement in industrial environments.
• Self-diagnostic function: Sensors should be able to output their own status (such as light source life, cleaning cycle, calibration coefficient) for predictive maintenance.
• On-site display and control: Optional touch controller for on-site calibration and data viewing, reducing the difficulty of upper computer integration.

4.3 Installation Layout Points

• Position selection: Try to choose a turbulent flow area that represents the overall water quality (e.g., after an elbow or tee), avoiding dead zones or areas where bubbles accumulate.
• Immersion depth: Ensure the probe is fully immersed and not directly facing bubble impact in the direction of water flow.
• Maintenance access: Reserve enough space for regular removal of the probe for calibration or cleaning. For self-cleaning sensors, periodic checks are still required.

V. Implementation Steps (from Planning to Commissioning)

The following are typical deployment steps for an online multi-parameter monitoring system in water treatment, which can be tailored based on project scale.

Step 1: Requirement Analysis and Parameter Confirmation

• Discuss with process, water quality laboratory, and management teams to clarify monitoring objectives (safety early warning, process control, discharge monitoring).
• List all parameters of concern and distinguish between parameters that "must be monitored online," "can be sampled offline," and "are only for laboratory analysis." For example, heavy metals generally still rely on laboratory testing.

Step 2: Sensor Selection and Procurement

• Determine the combination of online parameters based on the table above.
• Assess water quality conditions: temperature range, pressure, corrosiveness, scaling tendency, and select appropriate protection ratings (e.g., IP68) and materials.
• Obtain product certifications (e.g., CE, IP rating), technical manuals, and user cases from reliable suppliers.

Step 3: System Design

• Determine the installation method: immersion bracket, flow cell, or buoy type.
• Plan power supply and communication: 24VDC power or solar + battery; RS485 wiring to a control cabinet or remote terminal unit (RTU).
• Data acquisition and display: interface with existing SCADA, or build a standalone water quality monitoring platform.

Step 4: On-site Installation and Commissioning

• Install sensors according to the manufacturer's instructions, paying attention to cable shielding and grounding.
• Perform initial calibration: use standard buffer solutions (pH), zero turbidity water, conductivity standard solutions, etc., for multi-point calibration to ensure that the factory coefficients are suitable for the site water quality.
• Conduct a 72-hour continuous operation test, comparing with laboratory sampling data to verify consistency.

Step 5: Training and Establishment of Maintenance Plan

• Train operators on sensor working principles, cleaning cycles, calibration procedures, and troubleshooting.
• Develop a maintenance calendar: self-cleaning optical probes should be manually checked every 1-3 months (depending on water quality), electrode-type conductivity probes may require more frequent cleaning; conduct calibration verification every 3-6 months.

VI. Verification Methods and Data Quality Control

6.1 Offline Comparison Verification

Online sensor readings must be regularly compared with laboratory national standard methods to establish a correction relationship. Key steps:

• Use "simultaneous sampling" at the same sampling point—that is, at the moment of sensor reading recording, collect water samples and send them to the laboratory.
• Collect at least 20 paired data sets, calculate the correlation coefficient (R²) and mean deviation, allowing for a certain systematic error in sensor data, but trends must be consistent.
• For significant deviation points, analyze the causes (e.g., interfering substances, bubbles).

6.2 Online Alternative Calibration

For situations where the probe cannot be easily removed, the "standard substitution method" can be used:

• Turbidity: sealed solid standard plate (e.g., formazin equivalent standard).
• Organic matter: prepare known concentration standard solutions of potassium hydrogen phthalate (COD) or sucrose (TOC) and pass them through the sensor, recording reading consistency.
• Conductivity: use traceable standard conductivity solutions (e.g., 1413 μS/cm).

6.3 Data Review and Alarm Logic

• Set dead zones and alarm delays to prevent false alarms caused by bubbles or transient fluctuations.
• Establish combined alarms: for example, "turbidity > 2 NTU and lasting 10 minutes" triggers a raw water abnormality warning, while "turbidity > 5 NTU and TOC < normal value" may indicate an inorganic particulate event.

VII. Limitations: What Online Monitoring Can and Cannot Do

7.1 Advantages of Online Monitoring

• Real-time: second-level to minute-level data capture transient changes.
• Continuity: no manual sampling intervals, providing complete time series.
• Integrability: data directly feeds into automated decision-making.
• Cost amortization: lower long-term cost compared to intensive sampling.

7.2 Inherent Limitations of Online Monitoring

• Does not represent compliance: For regulatory indicators (e.g., 106 items for drinking water), the vast majority must still be tested in laboratories and produce accredited reports. Online data can serve as process control and early warning but cannot replace statutory testing.
• Interference and drift: Optical sensors may be affected by matrix effects at very high color or turbidity; conductivity sensors are affected by fouling and polarization effects, requiring regular calibration.
• Limited parameter coverage: For heavy metals, pesticides, specific organics, microorganisms, etc., existing online technologies are not mature enough or are costly, so laboratory verification is still required.
• Maintenance dependence: Even with self-cleaning features, manual intervention is still required under harsh water quality conditions, otherwise data reliability decreases.

7.3 Complementarity with Portable Testing

Portable multi-parameter test pens (e.g., Oromë Water Detective 4) can quickly screen 9 parameters including TDS, TOC, COD, turbidity, and conductivity, making them efficient tools for on-site inspection, water purification services, and household assessment. However, their design purpose is on-site screening and trend comparison; they cannot replace online fixed monitoring or provide real-time correlation of continuous data. Therefore, engineering practice often adopts a system of "fixed online monitoring as the primary method, supplemented by portable sampling."

8. Frequently Asked Questions (FAQ)

Q1: Is it normal that the value from an online TDS sensor differs greatly from laboratory results? A: Normal. Online TDS is usually converted from conductivity, with a fixed conversion factor (e.g., 0.5), but the actual factor varies with water quality. In addition, temperature compensation algorithms and electrode constant drift can introduce errors. Online TDS is mainly used for trend monitoring; the absolute value should be regularly compared with the drying method or a laboratory conductivity meter, and the sensor coefficient should be corrected accordingly.

Q2: Do multi-spectral sensors really require no reagents? How are they maintenance-free? A: Multi-spectral sensors measure the absorption spectrum of a water sample in the UV-visible range to estimate organic matter content, using physical optical methods without adding chemical reagents. However, reagent-free does not mean maintenance-free: the measurement window must remain clean, so sensors with automatic physical wipers or ultrasonic cleaning can greatly extend maintenance intervals, but periodic checks of cleaning effectiveness and optical window condition are still necessary.

Q3: Our factory discharge only needs COD to meet standards, why do we also need to monitor turbidity and color? A: COD is the core indicator for online organic matter monitoring, but turbidity reflects suspended solids concentration, and color reflects the sensory indicators of dissolved organics and dyes. The combination provides a more comprehensive understanding of discharge quality. Additionally, when the COD sensor needs calibration or when abnormal readings are suspected, the correlated changes in turbidity and color can provide corroborating evidence. Moreover, many local discharge standards require color and SS (suspended solids, related to turbidity), so online monitoring aids compliance.

Q4: Are multi-parameter sensors costly? How much more expensive than traditional multiple single probes? A: The purchase cost of a single multi-parameter probe is slightly higher than a simple single-parameter probe, but because it reduces installation accessories, cables, integration and commissioning work, and maintenance points, the total life-cycle cost is often lower. For example, NSDD6, which integrates 6 optical indicators, saves at least 50% of installation space and maintenance hours compared to purchasing a turbidity meter, COD meter, and colorimeter separately.

Q5: How often does an online monitoring system need calibration? A: There is no universal answer; it depends on water quality, sensor type, and cleaning measures. Generally, clean water (e.g., tap water) can be calibrated every 3-6 months; sewage or high-turbidity water may need calibration every 1-2 months. It is recommended to compare weekly during the initial phase and gradually determine the calibration cycle suitable for the plant's water quality.

Q6: Can online monitoring completely replace laboratory manual sampling? A: No! Regulations require laboratory-accredited reports, and online monitoring cannot cover all indicators (e.g., heavy metals, pathogenic bacteria). The role of online monitoring is real-time process control and exceedance early warning, combined with regular laboratory testing to form a complete water quality management system.

Conclusion

As one of the most popular water quality indicators, TDS indeed has irreplaceable industrial value in specific scenarios (e.g., pure water preparation). However, treating it as the sole yardstick for judging water safety is a technological cognitive regression. In fields such as process control, environmental emergency response, and water supply safety, the shift from "single TDS" to "multi-parameter collaborative monitoring" is an irreversible engineering trend.

Correctly building an online multi-parameter monitoring system requires moving beyond the rough mindset of "buy an instrument and install it." From clarifying monitoring needs, scientifically matching parameters, rational selection and integration, to thorough verification and maintenance, every step demands engineering rigor. When you find that water quality anomalies cannot be explained by TDS alone, look at turbidity, organic matter, color, and temperature. They build a three-dimensional water quality picture, which is the industrial language for ensuring water safety.