A cooling tower controller shows a free chlorine residual of 1.0 ppm. A technician takes a DPD sample at the same point and gets 0.6 ppm. Which one is right? The answer depends less on which instrument is more expensive and more on how each one measures chlorine, and what in that particular water can distort the reading.
There are four common ways to measure free chlorine in water treatment: DPD colorimetric analysis, membrane-type amperometric sensors, membrane-type potentiostatic sensors, and bare gold (membrane-free) sensors. Each has a different principle, a different set of weaknesses, and a different set of applications where it performs best.
The short answer is this. DPD is the reference method and the one to trust for verification, but it is a batch measurement that consumes reagents. Membrane sensors offer the best specificity among continuous sensors, because the membrane screens out most interferences, but the membrane fouls and needs maintenance. Potentiostatic membrane sensors are the more stable version of that design. Bare gold sensors need the least maintenance and respond fastest, but they respond to all oxidants, not only chlorine, and they depend on a controlled flow rate. No single technology is the most accurate in every situation.
Why free chlorine is harder to measure than it looks
When sodium hypochlorite is added to water, free chlorine exists in two forms: hypochlorous acid (HOCl) and the hypochlorite ion (OCl-). The balance between them depends on pH. At pH 6.5, most free chlorine is HOCl. At pH 8.5, most of it is OCl-. HOCl is the far stronger disinfectant, and it is also the form that most electrochemical sensors detect well.
This means any electrochemical sensor reading falls as pH rises, even when total free chlorine has not changed. Every continuous chlorine sensor therefore needs pH to be stable or compensated. DPD, by contrast, uses a buffered reagent that measures both forms together.
DPD colorimetric analysis
How it works. A reagent, N,N-diethyl-p-phenylenediamine, reacts with free chlorine to produce a magenta colour. A photometer measures the colour intensity, which is proportional to chlorine concentration. Handheld test kits and online analysers both use this principle; online analysers draw a sample, add reagent and read it in a repeating cycle every few minutes.
Strengths. DPD is the accepted reference method for regulatory testing and the method used to calibrate every electrochemical sensor. Because the reagent is buffered, it measures HOCl and OCl- together and is not affected by pH in the way sensors are. It is largely unaffected by flow rate and fouling.
Weaknesses. It is not continuous, which limits its use for responsive dosing control. Reagents degrade with age and temperature and must be replenished, and online analysers produce a small waste stream. Other oxidants can also produce a colour, including chlorine dioxide, ozone and oxidised manganese, and monochloramine can slowly appear in a free chlorine reading if the result is not read promptly.
Membrane-type amperometric sensors
How it works. A gold or platinum cathode and a silver anode sit in an electrolyte behind a thin membrane. HOCl diffuses through the membrane and is reduced at the cathode, producing a current proportional to its concentration. Many of these sensors are galvanic: the two dissimilar metals generate the polarising voltage themselves, without a separate reference electrode.
Strengths. The membrane provides physical selectivity. It allows small, uncharged molecules such as HOCl to pass, while blocking most dissolved ions, particles and many interfering substances. Because diffusion through the membrane controls how fast chlorine reaches the electrode, the reading is relatively insensitive to flow rate.
Weaknesses. The sensor sees mainly HOCl, so it is strongly pH-dependent and needs pH compensation above about pH 7.5. The membrane fouls with biofilm, oils and surfactants, slowing response and lowering the reading gradually, often without any obvious sign. The membrane and electrolyte need periodic replacement, and in two-electrode designs the anode is consumed over time, which shifts the calibration.
Membrane-type potentiostatic sensors
How it works. The membrane and electrolyte arrangement is the same, but the sensor uses three electrodes: a working electrode, a counter electrode and a reference electrode. An electronic potentiostat holds the working electrode at a precisely controlled potential relative to the reference, while the measuring current flows through the counter electrode.
Strengths. Because the reference electrode carries no current, the working potential stays stable as the electrolyte ages. The result is better long-term stability, better linearity and a lower detection limit than a two-electrode membrane sensor. It keeps the membrane’s selectivity advantage.
Weaknesses. It shares the membrane’s limitations: pH dependence, fouling, and periodic membrane and electrolyte replacement. It also costs more and needs more sophisticated electronics.
Bare gold (membrane-free) sensors
How it works. A gold working electrode is exposed directly to the water, usually in a potentiostatic arrangement with counter and reference electrodes. Oxidants in the water are reduced at the gold surface, generating a current. Pyxis Lab’s ST-765 and ST-766 series use this approach, combining the bare gold electrode with a pH electrode in the same sensor body so that pH compensation is built in.
Strengths. There is no membrane to foul or tear and no electrolyte to replenish, so routine maintenance is limited to cleaning and calibration. Response is fast, typically within seconds, which suits dosing control. The design tolerates surfactants and dirty water that would blind a membrane, and Pyxis quotes a typical electrode life of two years with a replaceable electrode head.
Weaknesses. Without a membrane, the electrode responds to every oxidant that can be reduced at its working potential. Pyxis states this openly in its datasheets, listing the sensors as non-selective and cross-sensitive to other oxidising species. The current also depends on how quickly fresh water reaches the electrode surface, so the reading changes with flow rate unless flow is held constant.
Two conditions that decide whether bare gold performs
A single oxidant in the water. If chlorine is the only oxidant present, cross-sensitivity has no practical effect. It becomes a problem when two or more oxidants coexist, because the sensor reports their combined current, weighted by how strongly each reacts at the electrode. Calibrating against DPD only holds while the ratio between oxidants stays constant. Typical problem cases are chlorine with monochloramine, which forms whenever ammonia or organic nitrogen is present, and systems that also use chlorine dioxide, ozone, hydrogen peroxide or peracetic acid.
Controlled flow. A bare electrode consumes chlorine at its surface, so the current depends on the thickness of the boundary layer of water around it. Faster flow gives a higher reading; slower flow gives a lower one. A flow reservoir with a rotameter, such as the Pyxis FR-50, holds flow constant, makes the flow rate visible, lets air escape before it reaches the electrode, and provides a convenient point for the DPD sample used in calibration. Installed directly in a line with varying flow, a bare gold sensor will report flow changes as chlorine changes.
How the four compare
| Criteria | DPD | Membrane amperometric | Membrane potentiostatic | Bare gold |
|---|---|---|---|---|
| Measurement | Batch, every few minutes | Continuous | Continuous | Continuous |
| Response | Minutes per cycle | Tens of seconds to minutes | Tens of seconds to minutes | Seconds |
| Selectivity | Moderate | Good | Good | Low, responds to all oxidants |
| pH dependence | Low (buffered reagent) | High | High | High, compensated in Pyxis designs |
| Flow dependence | Low | Low | Low | High, needs a flow reservoir |
| Fouling tolerance | Good | Poor with oils, surfactants, biofilm | Poor with oils, surfactants, biofilm | Better |
| Consumables | Reagents | Membrane, electrolyte | Membrane, electrolyte | Minimal |
| Best fit | Compliance, verification | Clean water, budget installations | Drinking water, high accuracy | Single-oxidant water with controlled flow and limited maintenance |
Accuracy on day one versus accuracy in month three
Laboratory accuracy and field accuracy are not the same thing. Under ideal conditions, DPD and a well-maintained potentiostatic membrane sensor will give the most specific free chlorine readings. In practice, most installations are serviced every few weeks rather than every few days, and each technology loses accuracy in its own way between visits.
A membrane sensor typically loses accuracy gradually as the membrane fouls, and the falling reading looks plausible enough that it often goes unnoticed. A DPD analyser loses accuracy when reagents age. A bare gold sensor has fewer parts that degrade, so in clean, single-oxidant water with a flow reservoir, it can hold its calibration as well as or better than a membrane sensor between maintenance visits.
Response time matters too. When a sensor drives a dosing pump, slow feedback makes the control loop overshoot and undershoot the setpoint, so the actual chlorine level swings more widely even if each individual reading is precise.
What this means for Malaysian applications
Malaysian municipal water is disinfected by chlorination with a free chlorine residual, so most industrial water drawn from the mains is a single-oxidant system. The main exception is ammonia: raw water ammonia events have repeatedly affected treatment plants in Selangor, and combined chlorine can appear when ammonia enters a system, whether from the supply or from process contamination.
A simple check during commissioning or a site survey is to measure DPD free and total chlorine on the same sample. If total is close to free, chloramines are not a concern and a single free chlorine sensor is appropriate. If there is a meaningful gap, the system needs a membrane sensor, or a combined free and total chlorine measurement such as the Pyxis ST-766SS-TFCL, which reports both so that the difference reveals combined chlorine.
Choosing the right method
Use DPD for regulatory reporting, for calibrating sensors, and for resolving any disagreement between instruments. Choose a membrane potentiostatic sensor for drinking water and other clean water where specificity matters and maintenance is reliable. Choose a bare gold sensor for single-oxidant water where low maintenance and fast response matter, provided it is installed with a flow reservoir and calibrated against DPD. Avoid relying on a bare gold sensor alone where chloramines or other oxidants are present alongside chlorine.
Back to the cooling tower at the start of this article: before assuming either reading is wrong, check the flow through the sensor, check whether combined chlorine is present, and check the age of the DPD reagent. One of the three usually explains the gap.
Autoflo Technology is the Malaysian distributor for Pyxis Lab chlorine sensors and analysers. For help choosing a free chlorine measurement for a specific water system, contact us at info@autoflotechnology.com.