An electromagnetic flowmeter, often shortened to magmeter or mag flow meter, measures the volumetric flow of an electrically conductive liquid using Faraday's law of electromagnetic induction. It has no moving parts and no obstruction in the bore, so it imposes negligible pressure drop and tolerates dirty, abrasive, and corrosive media that would foul or wear a mechanical meter. It is the dominant technology for water, wastewater, slurries, pulp, and conductive process fluids, but it cannot meter oils, hydrocarbons, gases, or ultrapure water because those media lack the conductivity the principle requires.
This page is a reference for procurement and design engineers specifying a magmeter for a $10K to $1M project. It covers how the meter works, where the technology splits into types, the field-excitation methods, the liner and electrode choices that decide service life, the spec-sheet numbers that drive selection, and the decision sequence that turns requirements into a model code.
Photo: ASDFS, CC BY-SA 3.0, via Wikimedia Commons
This guide is written for industrial purchasing engineers and design engineers. Across 6 chapters it covers the Faraday's-law principle, meter types and field excitation, liners and electrodes, media and material compatibility, key specification parameters, and the selection decision sequence, with 7 FAQs and manufacturer comparisons. Parameter ranges reference the public OIML R49 and ISO 4064 metrology standards for water meters, the IEC 60079 hazardous-area series, and the published technical datasheets of Endress+Hauser, Emerson Rosemount, ABB, Siemens, Yokogawa, and KROHNE.
Chapter 1 / 06
What is an Electromagnetic Flowmeter
An electromagnetic flowmeter measures the volumetric flow rate of a conductive liquid by exploiting Faraday's law of electromagnetic induction, the same law that governs electrical generators. Field coils mounted outside a lined, non-magnetic metering tube establish a magnetic field across the bore. When a conductive liquid flows through that field, the liquid behaves as a moving conductor and a voltage is induced perpendicular to both the field and the direction of flow. Two electrodes set flush with the liner, on a diameter at right angles to the field, sense that voltage. The induced voltage is directly proportional to the average flow velocity, the magnetic flux density, and the distance between the electrodes, which is the pipe inner diameter.
Because the transmitter holds the field strength and the bore constant, the measured electrode voltage is a clean linear function of average velocity, and multiplying velocity by the known cross-sectional area gives volumetric flow. The decisive engineering consequence is that the reading does not depend on the fluid density, viscosity, temperature, pressure, or flow profile turbulence in the way that differential-pressure, turbine, and variable-area meters do. A magmeter calibrated on water reads the same on slurry of the same velocity, which is why it dominates water and wastewater metering.
The principle was described by Michael Faraday in 1832, who attempted, unsuccessfully, to measure the flow of the River Thames from the voltage induced by the Earth's magnetic field. The first practical industrial electromagnetic flowmeters appeared in the 1950s once stable field excitation and high-impedance amplifiers were available. Early commercial meters used alternating-current field excitation; the pulsed-direct-current excitation that defines most modern meters was introduced in the 1970s to suppress the zero drift that plagued AC designs. Today the magmeter is one of the most widely installed flow instruments in the world, spanning bores from a few millimetres in dosing skids to over two metres in municipal water mains and hydroelectric intakes.
The hard boundary of the technology is conductivity. The induced signal is a few millivolts at best, and the source impedance seen by the amplifier rises as fluid conductivity falls. Below the meter's rated minimum, commonly around 5 microsiemens per centimetre for general liquids and about 20 microsiemens per centimetre for demineralised water, the signal sinks into electrode and amplifier noise and the reading becomes unstable or zero. For that reason a magmeter cannot meter ultrapure water, deionised water, distilled water, hydrocarbons, fuels, lubricating oils, or any gas. Those services belong to Coriolis, ultrasonic, vortex, or thermal-mass meters. Recognising this limit at the requirements stage prevents the single most common specification error with this instrument.
Four engineering attributes determine whether a magmeter will serve well over a decade: liner and electrode compatibility with the medium, the accuracy and turndown grade, the field-excitation scheme matched to process noise, and the quality of grounding and installation. The chapters that follow take each of these in turn, because the meter body itself rarely fails. Failures in the field almost always trace to a liner attacked by the medium, electrodes coated by grease or scale, a meter running part-full, or a missing earth path.
Chapter 2 / 06
Meter Types and Construction
Electromagnetic flowmeters divide along three construction axes: how the transmitter is mounted, how the sensor connects to the pipe, and whether the sensor fills the bore or is inserted into it. These choices do not change the underlying physics, but they decide installed cost, accuracy grade, and serviceability. The table below summarises the main families and where each fits.
Type
Construction
Typical Size Range
Best Fit
Full-bore flanged
Sensor flanged into the line, transmitter compact or remote
Single-point probe through a tapping in a large pipe
DN200 to DN3000+
Large mains where full-bore is uneconomic
Partially filled / open channel
Special electrode array for non-full pipes
DN200 to DN2000
Gravity sewers, drainage
Compact versus remote transmitter. In a compact (integral) meter the transmitter sits directly on the sensor, which simplifies wiring and is preferred for accessible, moderate-temperature installations. In a remote (separate) design the transmitter is wall-mounted up to tens of metres away and connected by a dedicated signal-and-coil cable. Remote mounting is chosen when the sensor is buried, submerged, in a hot or vibrating location, or otherwise hard to reach. The signal cable must be the manufacturer's specified low-noise type, because the millivolt electrode signal is easily corrupted.
Full-bore versus insertion. A full-bore meter measures the entire flow profile and is the accurate, traceable choice; its cost rises steeply with diameter because the liner, coils, and body all scale. An insertion or probe meter measures velocity at one point and infers the average from an assumed profile, which trades accuracy for a far lower price on very large mains. Insertion meters are typically specified at a few percent of reading rather than the half-percent of a full-bore meter, and they demand long, well-conditioned straight runs to keep the profile predictable.
Two-wire versus four-wire power. Classic magmeters are four-wire devices: power feeds the field coils separately from the 4-20 mA signal loop, giving a strong field and a wide turndown. Loop-powered two-wire magmeters, which energise both coils and output from the 4-20 mA loop, are now offered up to roughly DN200 (8 inch) and simplify wiring to a DCS, at the cost of a weaker field and reduced low-flow performance. For municipal and battery or solar metering, dedicated low-power and battery-operated variants run the field in short pulses and have cut typical power draw from several hundred watts in old AC designs to roughly 10 to 15 watts.
Construction also determines pressure rating and connection. Flanged sensors follow the pipe flange standard, commonly PN10 to PN40 in Europe or ASME Class 150 to Class 300 in North America, with higher ratings available for specific bores. Sanitary meters use hygienic clamps and electropolished wetted surfaces with the liner flush to the bore so there is no crevice to trap product. Selecting the construction family is the first branch of the decision tree in Chapter 6, because it constrains every later choice.
Chapter 3 / 06
Field Excitation Methods
How the coils are driven, the field-excitation method, is the single biggest factor in a magmeter's zero stability and its tolerance of process noise. Three schemes dominate: alternating-current excitation, low-frequency pulsed-direct-current excitation, and dual-frequency excitation that blends the two. The table compares their engineering behaviour.
Excitation
Field Frequency
Zero Stability
Noise Immunity
Best Fit
AC (mains)
50 / 60 Hz
Poor (drift)
High
Noisy slurry, pulp, high solids
Pulsed DC (square wave)
~3 to 30 Hz
Excellent
Moderate
Clean water, general process
Dual-frequency
Low + high combined
Excellent
High
Noisy media needing stable zero
AC excitation drives the coils with mains-frequency current. The continuous strong field gives a high signal-to-noise ratio, which historically made AC meters robust on electrically noisy media such as mining slurry and paper pulp. The drawback is zero stability. The changing field induces a transformer voltage in the signal loop, electrode contamination shifts the baseline, and mains pollution moves the zero crossing, so AC meters need periodic manual re-zeroing and draw substantial power. AC excitation is now reserved for difficult high-noise services where its raw signal strength still wins.
Pulsed-DC excitation switches the field on and off as a low-frequency square wave, typically a small fraction of line frequency. The key advantage is that the transmitter samples the electrode voltage while the field is off, capturing the true zero, and subtracts it from the field-on reading. This continuously cancels zero drift, electrode polarisation, and process-generated DC offsets without operator intervention, and it uses far less power than AC. The trade-off is lower bandwidth, so a pure pulsed-DC meter can struggle with very noisy slurries. Pulsed DC is the default for potable water, wastewater, and clean process liquids, and it is what most modern catalogue meters use.
Dual-frequency excitation superimposes a low-frequency component for stable zero on a higher-frequency component for noise rejection, so the meter keeps the automatic zero correction of pulsed DC while gaining the slurry tolerance once unique to AC. It is the premium option offered on flagship process meters where the medium is both noisy and demands a rock-steady zero, for example in mineral processing and high-solids effluent. It costs more in electronics and power than a basic pulsed-DC meter.
Two installation details interact strongly with excitation. First, grounding: the fluid and the meter body must share a defined earth reference so the electrode signal is measured against a stable baseline. Where the adjacent pipe is plastic, lined, or coated, grounding rings or grounding electrodes carry the reference, and a missing earth path corrupts the reading more severely than any straight-run shortfall. Second, the meter must run completely full at the electrode plane; modern meters include empty-pipe detection that watches electrode impedance and flags a part-full or empty bore rather than reporting a false flow.
Chapter 4 / 06
Liners, Electrodes and Media
Two wetted components decide whether a magmeter survives the medium: the liner that insulates the metering tube and the electrodes that contact the fluid. The metering tube must be non-magnetic and electrically insulated from the fluid, so a dielectric liner is bonded inside it. The electrodes must resist the same chemistry as the liner without coating or corroding. Getting both right for the specific medium is the difference between a ten-year service life and a failure in months.
Liner selection balances temperature, chemical attack, and abrasion. Hard rubber is the low-cost choice for ambient water and neutral fluids, rated to roughly +80 degrees Celsius. Polyurethane gives the best abrasion resistance for wastewater, sand, grit, and mild slurries, but it tolerates only a narrow band near -20 to +50 degrees Celsius. PTFE (polytetrafluoroethylene) resists almost all chemicals, acids, and bases and serves to roughly +130 to +180 degrees Celsius depending on the maker, but it is mechanically soft and can cold-flow or abrade under high velocity. PFA (perfluoroalkoxy) is a tougher fluoropolymer chosen for vacuum service and demanding chemical-plus-temperature duty. Ceramic (alumina) is the hardest liner, handling severe abrasion and temperatures above +250 degrees Celsius, but it is brittle, intolerant of thermal shock, and limited in line size.
Liner
Max Temperature
Strength
Avoid
Hard rubber
~ +80 °C
Low cost, neutral water
Hot or aggressive chemicals
Polyurethane
~ -20 to +50 °C
Best abrasion, slurry / grit
High temperature, solvents
PTFE
~ +130 to +180 °C
Broad chemical resistance
High-velocity abrasion, vacuum
PFA
~ +150 to +180 °C
Chemical + vacuum + temperature
Heavy abrasive solids
Ceramic (Al2O3)
> +250 °C
Extreme abrasion and heat
Thermal shock, large bores
Electrode selection follows the same media logic as a wetted diaphragm. 316L stainless steel (1.4435) is the default for potable water, industrial water, municipal wastewater, and weakly corrosive organic acids. Hastelloy C (Alloy C22 / C276) is the chemical-industry standard against hydrochloric acid, wet chlorine, and chlorides. Titanium suits seawater, brine, chlorides, hypochlorite, and oxidising acids. Tantalum handles concentrated acids and aggressive halogens. Platinum and platinum-iridium alloys add heat resistance for high-temperature service. Where coating is a concern, bullet-nose or removable electrodes, and on some designs ultrasonic or mechanical cleaning electrodes, keep the surface clear of grease and scale.
Conductivity and media limits govern whether the meter can be used at all. The medium must conduct, with a typical minimum near 5 microsiemens per centimetre for general liquids and about 20 microsiemens per centimetre for demineralised water; specialised high-input-impedance transmitters extend this down toward 0.05 to 0.1 microsiemens per centimetre. The meter cannot read ultrapure water, deionised water, hydrocarbons, oils, fuels, or gases. The liner and electrodes set the temperature and pressure envelope, and abrasive solids dictate the liner hardness. The lookup below pairs common media with a starting material choice; always confirm against the manufacturer corrosion chart for actual concentration, temperature, and velocity before purchase.
Medium
Suggested Liner
Suggested Electrode
Potable / industrial water
Hard rubber or PTFE
316L stainless steel
Municipal wastewater, sludge
Polyurethane or hard rubber
316L or Hastelloy C
Abrasive slurry, sand, mining
Polyurethane or ceramic
Hastelloy C or tantalum
Dilute acids and bases
PTFE or PFA
Hastelloy C or tantalum
Seawater, chlorides, hypochlorite
PTFE or hard rubber
Titanium
Food, beverage, pharma CIP
PFA or PTFE (polished)
316L electropolished
Chapter 5 / 06
Key Specification Parameters
A magmeter datasheet can list two or three dozen lines, but only a handful drive selection. The decisive parameters are accuracy and its reference basis, repeatability, turndown and velocity range, the minimum conductivity, the liner and electrode temperature limits, the process pressure rating, the output and communication options, and the ingress and hazardous-area ratings. Each is explained below so the numbers can be compared like-for-like across manufacturers.
Accuracy and reference basis. Magmeters are normally specified as a percent of reading (percent of rate), not percent of full scale, which is why they hold accuracy across a wide turndown. A typical process meter is rated about plus or minus 0.5 percent of rate plus a small velocity floor such as plus or minus 1 to 2 millimetres per second; premium options reach plus or minus 0.2 percent of rate, and laboratory-grade meters approach plus or minus 0.1 to 0.15 percent. The velocity floor matters at low flow because it sets a fixed error that grows as a fraction of reading near the cutoff. Always check whether a quoted figure is percent of rate or percent of full scale before comparing two meters.
Turndown, velocity range, and low-flow cutoff. The usable velocity span runs roughly 0.1 to 10 metres per second, and turndown is commonly 20:1 and can reach about 100:1 on high-end meters. The design target is to size the bore so maximum flow gives about 2 to 3 metres per second, which keeps the signal strong, flushes the electrodes, and limits liner wear. Below the configured low-flow cutoff, often a few millimetres per second of velocity, the transmitter forces the reading to zero to suppress noise and residual offset.
Conductivity, temperature, and pressure. Minimum conductivity is a hard limit, near 5 microsiemens per centimetre for general liquids and about 20 for demineralised water. The liner and electrode set the medium and ambient temperature limits, summarised in Chapter 4; PTFE and ceramic reach the highest medium temperatures while rubber and polyurethane are constrained. Pressure rating follows the flange or wafer standard, commonly PN10 to PN40 or ASME Class 150 to 300, with higher ratings for specific bores.
Output signal and communication is the interface to the control system, which often closes a flow-control loop back to a control valve. The mainstream options are:
4-20 mA + HART: the default analog loop with digital configuration and diagnostics overlaid, used on most installations.
Pulse / frequency and digital output: scaled volume pulses for totalisers, batch controllers, and flow computers.
Modbus RTU / RS-485: simple digital integration for SCADA and metering networks.
PROFIBUS PA / FOUNDATION Fieldbus: shared-bus digital communication for large DCS projects.
Industrial Ethernet (PROFINET, EtherNet/IP, Modbus TCP): the smart-factory trend for centralised data and diagnostics.
Ingress protection and approvals. Housing protection commonly ranges IP66 / IP67 for outdoor and washdown duty up to IP68 for buried or submerged sensors, the latter essential for direct-burial water meters. Hazardous-area approvals follow IEC 60079 (ATEX, IECEx, plus NEPSI in China and FM or CSA in North America) for installations in flammable atmospheres. For billing and custody transfer the meter must carry OIML R49 and ISO 4064 pattern approval to Class 1 or Class 2, which define the permissible error across the flow range from the minimum flow Q1 to the overload flow Q4. Sanitary service adds 3-A or EHEDG hygienic certification.
Chapter 6 / 06
Selection Decision Factors
To turn the preceding five chapters into a model code, work the decision sequence below in order. Most selection errors are not a single wrong number but a decision made at the wrong level, for example fixing the line size before checking the target velocity. The eight steps can serve as a fixed RFQ template.
Confirm the medium is meterable: verify the fluid is conductive at or above the meter's minimum, near 5 microsiemens per centimetre (about 20 for demineralised water). If it is oil, fuel, gas, or ultrapure water, stop here and choose Coriolis, ultrasonic, vortex, or thermal-mass instead.
Size for velocity, not for the pipe: target about 2 to 3 metres per second at maximum flow. If the line size gives the wrong velocity, fit a reducer rather than oversize the meter, so it does not sit in its low-signal region.
Choose the construction family: full-bore flanged, wafer, sanitary, or insertion (Chapter 2). Full-bore for accuracy and custody transfer; insertion only where a full-bore meter is uneconomic on a very large main.
Match liner and electrode to the medium: use the Chapter 4 lookup for temperature, chemistry, and abrasion, then confirm against the manufacturer corrosion chart for actual concentration, temperature, and velocity.
Pick the field excitation for the noise level: pulsed DC for clean water and process service; AC or dual-frequency for noisy slurries, pulp, and high-solids media where zero stability and noise rejection are both needed.
Set accuracy grade and approvals: plus or minus 0.5 percent of rate for general control, plus or minus 0.2 percent or better for demanding duty, and OIML R49 or ISO 4064 Class 1 or Class 2 for billing. Add ATEX / IECEx / NEPSI for hazardous areas and 3-A / EHEDG for hygienic service.
Define power, output, and protection: four-wire for widest turndown, two-wire loop-powered for simple wiring up to about DN200, or battery and low-power for remote municipal meters. Select 4-20 mA HART, pulse, Modbus, fieldbus, or Ethernet, and IP66 / IP67 for outdoor, IP68 for buried or submerged.
Plan installation and grounding: allow about 5 diameters upstream and 2 downstream of straight run, keep the meter full at the electrode plane, and specify grounding rings or electrodes where the adjacent pipe is plastic, lined, or coated. Bad grounding corrupts the signal more than a short straight run.
One last dimension is manufacturer serviceability: local calibration capability, spare liner and electrode availability, in-situ verification or self-diagnostic features, and DD or driver files registered with the relevant fieldbus body. These seem secondary at purchase but decide repair response after years of service. Endress+Hauser (Proline Promag W and P), Emerson Rosemount (8750W, 8732), ABB (WaterMaster, ProcessMaster), Siemens (SITRANS F M MAG 5100 W with MAG 5000 or 6000 transmitter), Yokogawa (ADMAG TI and AXG), and KROHNE (OPTIFLUX) all maintain calibration and service infrastructure in China and globally, which makes them reliable for large projects. Regional suppliers such as Sino-Inst, Supmea, and other Chinese makers offer compatible meters with NEPSI and CPA pattern approval at lower cost for non-critical and municipal loops.
FAQ
What is the minimum conductivity an electromagnetic flowmeter can measure?
Most industrial electromagnetic flowmeters require a minimum liquid conductivity of about 5 microsiemens per centimetre, with demineralised water typically needing 20 microsiemens per centimetre because its dielectric behaviour raises the source impedance. High-resolution transmitters with high-input-impedance front ends can reach 0.05 to 0.1 microsiemens per centimetre. Below the rated threshold the induced signal collapses into electrode and amplifier noise, giving unstable or zero readings. This is the hard physical limit of the technology: it cannot meter ultrapure water, deionised water, hydrocarbons, oils, or any non-conductive fluid. For those services choose Coriolis, ultrasonic, or vortex meters instead.
How does an electromagnetic flowmeter work?
It applies Faraday's law of electromagnetic induction. Field coils outside the lined metering tube create a magnetic field across the bore. As a conductive liquid flows through that field it acts as a moving conductor and induces a voltage proportional to the average flow velocity, the magnetic flux density, and the pipe inner diameter. Two electrodes set flush in the liner, perpendicular to both the field and the flow, pick up that millivolt-level voltage. The transmitter holds the field constant and computes volumetric flow from the measured voltage. Because the signal depends only on velocity and bore, the reading is independent of fluid density, viscosity, temperature, pressure, and Reynolds number, which is why magmeters are so widely used on water and slurries.
What is the difference between AC and pulsed-DC field excitation?
AC excitation drives the coils with mains-frequency current, giving a strong signal and a high signal-to-noise ratio that resists slurry and pulp noise, but it suffers zero drift from transformer pickup, electrode contamination, and mains pollution, and it draws more power. Pulsed-DC (square-wave) excitation switches the field on and off at typically 1/8 to 1/2 of line frequency and samples the zero level during the off period, so the transmitter continuously subtracts zero shift and electrode-polarisation error without manual re-zeroing, at lower power. Pulsed DC is the default for clean water and process service. AC or higher-frequency dual-frequency excitation is preferred for noisy slurries, pulps, and high-solids media.
How much straight pipe run does a magnetic flowmeter need?
The common industry guideline is five pipe diameters of straight run upstream of the electrode plane and two diameters downstream, which lets the velocity profile recover symmetry after fittings. Severe disturbances such as two elbows in different planes, partly open valves, or pumps may call for ten diameters or more upstream; some manufacturers reduce the requirement to three diameters upstream and two downstream for their specific designs. The meter must run full at the electrode plane, so install it in a rising line or a low point and never at a high point where gas can collect. Good grounding to the fluid matters more than straight run: a missing earth path corrupts the signal even with perfect piping.
How do I choose the liner material?
Match the liner to temperature, chemistry, and abrasion. Hard rubber suits ambient water and neutral fluids to roughly +80 degrees Celsius at low cost. Polyurethane gives the best abrasion resistance for wastewater, sand, and mild slurries but a narrow temperature band near -20 to +50 degrees Celsius. PTFE handles aggressive chemicals, acids, and bases and tolerates about -20 to +130 to +180 degrees Celsius depending on the maker, but it is soft and can cold-flow or abrade under high velocity. PFA is a more durable fluoropolymer for vacuum and demanding chemical-and-temperature duty. Ceramic (alumina) is the hardest option, resisting strong abrasion and temperatures above +250 degrees Celsius, but it is brittle, cannot take thermal shock, and limits line size.
What accuracy can an electromagnetic flowmeter achieve?
Standard process magmeters are specified as a percent of reading rather than percent of full scale, typically plus or minus 0.5 percent of rate plus a small velocity floor such as plus or minus 1 to 2 millimetres per second. High-end options reach plus or minus 0.2 percent of rate, and the best laboratory-grade meters approach plus or minus 0.1 to 0.15 percent. Because the error is a percent of reading, accuracy stays good across a wide turndown, commonly 20:1 and up to about 100:1 down to the low-flow cutoff. For billing and custody transfer the relevant benchmark is OIML R49 and ISO 4064, where Class 1 and Class 2 define the permissible error envelope across the flow range from Q1 to Q4.
What flow velocity range should I design for?
Electromagnetic flowmeters work over roughly 0.1 to 10 metres per second, but the practical design target is to size the bore so that maximum flow gives about 2 to 3 metres per second. That keeps the signal strong, flushes the electrodes to limit coating, and limits liner wear. Running too slow weakens the signal and lets solids settle or grease build up on the electrodes; running consistently above about 5 metres per second accelerates abrasion of soft liners. When the process flow does not match the line size, fit a reducer so the meter sees the target velocity rather than oversizing the meter to the pipe, since an oversized meter sits in its low-signal region most of the time.