Specifying a conductivity meter for a chemical dosing skid is a chemistry-first decision: the sensor family, cell constant, wetted material, and signal output have to align with the specific reagent, concentration window, and DCS interface before any vendor shortlist is built.
Dosing skids feed coagulants, flocculants, disinfectants, acids, and caustics into water or process streams, and conductivity is the proxy variable that tracks dissolved ion load, contamination drift, and dilution of the dosate [S1]. Get the meter wrong and the dosing loop either chases noise, coats over with process film, or hands the controller a signal it cannot scale into a stroke setpoint.
Why conductivity is the workhorse measurement on a dosing skid
Electrical conductivity in water tracks the concentration of dissolved ionic species, and dosing skids use that signal both as a quality flag for the incoming stream and as a feedback variable for reagent trim [S1]. In coagulant and flocculant service the meter detects shifts in dissolved solids that change demand; in sodium hypochlorite, acid, and caustic feed it confirms the bulk dosing stream is at the expected strength and flags dilution events upstream of the injection point [S1, S3].
Continuous monitoring converts the skid from a fixed-rate feeder into a closed trim loop, which is why the conductivity sensor is treated as a control-grade instrument, not a lab-style probe [S1]. On skids feeding potable water the conductivity loop is part of the regulatory record, so its signal integrity is treated with the same discipline as flow and pH.
Sensor family: contacting two-electrode vs four-electrode vs inductive (toroidal)
Three conductivity sensor families cover the dosing envelope, and the boundary between them is set by conductivity range, fouling tendency, and chemical attack risk [S2, S3]. [S1]
Contacting two-electrode cells with a low cell constant (k = 0.1 cm⁻¹) cover 0.1 to 200 µS/cm service and pure-water applications, but the electrodes are wetted and exposed to coating [S2]. Four-electrode cells push the upper end toward the 10,000 µS/cm industrial-water band and tolerate moderate fouling because the measurement current is separated from the sense electrodes [S3]. Inductive (toroidal) sensors such as the Endress+Hauser Indumax CLS50/CLS50D eliminate electrode contact entirely, use a PFA body for chemical resistance, and offer a PEEK variant rated to 180 °C; the C750 toroidal is built specifically for sodium hypochlorite, acids, and caustics where contacting sensors scale or fail [S2, S3].
For most chemical dosing skids — sodium hypochlorite, sulfuric acid, sodium hydroxide, ferric chloride — the inductive sensor is the default pick; two-electrode cells remain appropriate only on the low-conductivity water side upstream of the chemical injection point [S2, S3].
Range, temperature, and pressure envelopes
Industrial conductivity ranges span four orders of magnitude, and the sensor's cell constant, body rating, and seal package have to be matched to the worst-case operating point, not the nominal point [S3].
Typical windows: ultra-pure water 0–20 µS/cm, general industrial water 0–10,000 µS/cm, and chemical solutions reaching roughly 50,000 µS/cm at the high end [S3]. A pure-water inductive sensor with k = 0.1 cm⁻¹ is rated 0.1 to 200 µS/cm, –10 to 60 °C, with pressure limits of 7 bar abs at 20 °C derating to 1 bar abs at 60 °C, and the PEEK-bodied inductive variant extends the thermal ceiling to 180 °C for hot chemical service [S2]. Specifying a meter only against the design point — without confirming the derate curve — is the most common cause of premature seal failure on dosing skids.
Wetted materials and chemical compatibility
Wetted material selection on a conductivity meter is governed by the same logic as a flow meter or industrial valve on the same skid: PFA, PEEK, PTFE, and 316L stainless each have a defined envelope, and crossings outside that envelope drive failure modes that mimic calibration drift [S2, S7]. [S2]
PFA-bodied inductive sensors handle most mineral acids, bases, and sodium hypochlorite at ambient to moderate temperatures [S2, S3]. PEEK raises the ceiling to roughly 180 °C and is the preferred body for hot caustic and hot acid service [S2]. Stainless-electrode contacting cells are restricted to clean, non-oxidizing streams because halides, oxidizers, and reducing agents attack the metal and shift the cell constant [S3]. The C750-class toroidal sensor is framed specifically for harsh chemicals where contacting electrodes scale, pit, or corrode [S3].
Signal output and skid integration: 4–20 mA, HART, and the DCS
The meter's output has to land on a controller that already exists on the skid, and on most chemical dosing skids that controller is a PLC feeding a plant DCS via 4–20 mA analog or HART [S9].
Stroke positioners on metering pumps are specified to accept 4–20 mA DC signals, and the pump LVDT feedback is typically 24 V DC two-wire, so the conductivity transmitter needs a powered 4–20 mA output and, where plant practice allows, HART for remote range, calibration, and diagnostic access from the DCS [S9]. A dedicated conductivity loop is usually wired to the DDCMIS or local panel for the operator, with a parallel signal to the plant DCS, and the auto stroke controller is fed from the DCS or local panel, never from a hand-held calibrator in normal operation [S9]. Where the skid carries multiple pumps, the meter should have a second output or a multiplexer so duty and standby pumps can be slaved to the same feedback without a manual swap.
Skid-level criteria: duty-standby, flow pairing, and packaging
Process tolerance on potable-water and continuous chemical dosing skids is often non-negotiable: a single point of failure on the conductivity loop is not acceptable, and that pushes the spec toward duty-standby architecture [S4].
Pre-assembled dosing packages minimize installation time and process interruption, and most skid builders pair the conductivity meter with a redundant sample line, isolation valves, and a pressure transmitter for proof-of-flow interlocks [S4, S10]. A chemical metering skid typically bundles pumps, flowmeters, valves, a control panel, and the conductivity loop into a single engineered package with a small footprint relative to field-built equivalents [S10]. Parallel metering streams are the standard duty-standby approach; a single conductivity meter can serve both streams if it sits in a common sample manifold with a valve-per-stream switchover, but two meters on separate manifolds are the safer pattern on critical service [S4].
Failure modes and what to verify before sign-off
Most conductivity meter problems on dosing skids are not electronics failures — they are coating, polarization, and ground-loop artifacts that look like calibration drift to a technician [S2, S3]. [S3]
Coating of two-electrode cells in high-fouling streams shifts the cell constant and reads low; inductive sensors tolerate coating better because the measurement is non-contact, which is why they are preferred in hypochlorite, lime slurry, and ferric chloride service [S2, S3]. Polarization errors dominate at high conductivity and small cell constants, so specifying k = 0.1 cm⁻¹ sensors above 200 µS/cm is a known mistake [S2]. Ground loops between the sensor, the pressure transmitter, and the DCS can inject 50/60 Hz noise; verifying isolated outputs and a single-point ground before commissioning is faster than chasing a "bad calibration" for a week [S9].
Trackable signals: dosing skid builders are publishing more detailed material traceability on PFA and PEEK wetted parts through 2026 Q2, and plant DCS upgrades are pulling HART diagnostics from conductivity transmitters into asset-management dashboards — both worth confirming with the skid OEM before the next RFQ goes out [S3, S9].