Level transmitters convert tank level into a continuous output, most commonly 4–20 mA, HART, or RS485/Modbus RTU, and the right pick is governed by media, tank geometry, and control-system protocol rather than brand [S1][S2].
When the application calls for continuous monitoring, inventory reconciliation, or closed-loop control, specify a transmitter; when only single-point high/low alarms or pump protection are needed, a level switch is the lower-cost path and the transmitter vs switch decision becomes a question of whether you need a real number or a binary state [S2][S3].
Match Sensing Principle to Tank and Media
Hydrostatic differential pressure transmitters calculate level from the head pressure at the tank bottom and are the conventional fit for closed, pressurized vessels, but they require a known, stable specific gravity and compensating leg for sealed tanks [S4].
Capacitance level transmitters read level by measuring the change in capacitance between electrodes as the dielectric changes, suiting conductive and non-conductive liquids alike, including foams and slurries where ultrasonic struggles [S1].
Magnetostrictive level transmitters resolve level to ±1 mm via a waveguide and a float, the standard for custody-transfer and small-vessel high-accuracy service, but they need a moving float that limits use with heavy sludge or high-viscosity media [S2].
Guided-wave radar and free-space radar (typically 6 GHz, 26 GHz, and 80 GHz variants) are non-contact and unaffected by foam, vapor, or turbulence, which is why they are widely specified for hydrocarbon, chemical, and steam-boiler drums where hydrostatic drift would otherwise be unacceptable [S1].
Output, Protocol, and Cabling
4–20 mA analog remains the workhorse because live-zero (4 mA ≠ 0) gives a hardware fault signal and a single twisted pair can run hundreds of metres without repeaters; HART overlays digital diagnostics on the same two wires for asset management, while Foundation Fieldbus and PROFIBUS PA are fully digital protocols and are not compatible with the HART signal-on-4–20 mA scheme [S1].
RS485/Modbus RTU is the practical choice for multi-tank SCADA networks and for sites where the cable is already a serial bus; cellular telemetry and LoRa wireless serve remote or unmanned tanks where running cable is uneconomic [S1].
Confirm the PLC/DCS input card before you order: a 4–20 mA input cannot read an RS485 bus, and a HART-only maintenance tool will be blind to a PROFIBUS PA device, so the pressure transmitter family of integration rules (output type, power supply, cable distance, control-system input type, protocol, and alarm mapping) applies one-to-one to level instruments [S1].
Accuracy, Range, and Temperature Limits

Level transmitters carry a defined accuracy spec expressed as a percentage of the calibrated span (commonly ±0.1% to ±0.5% of span for industrial units, tighter for custody-transfer), and because accuracy scales with span, specifying a 0–10 m unit on a 1 m tank wastes the instrument's resolution [S1][S3].
Process temperature drives the wetted-material choice: 316L stainless is fine for water and most chemicals below 200 °C, while PTFE or PVDF liners extend service into aggressive acids and higher temperatures, and process connection ratings (typically 150# / 300# / 600# flange classes) must exceed the maximum working pressure with margin [S1].
For sealed or pressurized vessels, the temperature transmitter on the reference leg must be paired with the level device so that density compensation stays in the loop, otherwise a 10 °C shift changes the indicated level by roughly 0.2% on water service and far more on hydrocarbon service [S4].
Material Compatibility and Hazardous Area
Wetted parts (probe, gasket, housing) determine service life more than the datasheet accuracy figure; a transmitter specced to ±0.075% of span will still fail early if the probe is 304 stainless in a 10% HCl tank, which is why corrosive media call for Hastelloy, PVDF, or PTFE-lined construction [S1][S4].
Hazardous-area classification dictates the certification path: ATEX 2014/34/EU and IECEx govern Zone 0/1/2 and Division 1/2 locations, and the cert package (Ex ia, Ex d, Ex db eb) must cover both the process media group and the ambient gas group, not just one [S4].
Material compatibility and overpressure limits routinely outweigh baseline accuracy in total cost of ownership, because an unscheduled shutdown caused by sensor drift or corrosion costs more in lost production than the price difference between a standard and a high-stability probe [S4].
Comparison: Five Common Principles on Buyer Criteria

Five sensing principles line up against the criteria that drive most purchase decisions: hydrostatic DP, capacitance, magnetostrictive, guided-wave radar, and ultrasonic each answer a different question about the tank [S1][S2][S4].
Cost, lowest to highest: hydrostatic DP and ultrasonic are the budget tier, guided-wave radar sits in the mid range, and magnetostrictive and high-frequency radar are the premium tier. Accuracy: magnetostrictive leads at roughly ±1 mm, hydrostatic and guided-wave radar run ±0.1–0.5% of span, capacitance and ultrasonic are typically ±0.5–1% of span. Foam, vapor, and turbulence tolerance: radar and guided-wave radar are best, hydrostatic DP and capacitance are workable with calibration, ultrasonic is the worst because the airborne sound path is disrupted by foam. Moving parts: magnetostrictive has a float, the rest are static, which is why magnetostrictive is excluded from heavy-sludge service. Installation: hydrostatic DP needs a wet leg or remote seal, radar and guided-wave radar need a stilling well or process connection above the maximum level, ultrasonic needs line-of-sight to the liquid surface [S1][S2][S4].
If a buyer is choosing for a clean, open atmospheric water tank under 10 m, hydrostatic or ultrasonic is the economic answer; for a pressurized hydrocarbon separator with foam, guided-wave radar is the engineering answer; for a small additive day-tank that needs ±1 mm and a hardwired analog signal into a PLC, magnetostrictive is the engineering answer despite its higher unit cost [S1][S2].
Who Should NOT Choose the Mainstream Hydrostatic DP
Do not specify a hydrostatic differential pressure transmitter when the media specific gravity varies with temperature or composition, because every 1% change in density shows up as a 1% level error that no zero trim will fix [S4].
Avoid hydrostatic DP on tall vessels with a single tap at the bottom when the tank is open to atmosphere and large temperature swings are expected, because the reference leg will condensate, drift, and force weekly calibration walks; a non-contact radar or guided-wave radar is the correct substitution in that case [S1][S4].
Hydrostatic DP is also a poor fit for vessels with aggressive internal mixing or vortex breakers near the tap, where local pressure fluctuations alias into level noise, and for any application that requires sanitary CIP/SIP, where the wet leg, oil fill, and crevices are impossible to clean to 3-A or EHEDG standards [S1].
Installation, Commissioning, and Shortlist Logic

Confirm four things before sign-off: the calibrated span must be sized to the working level (not the full tank), the process connection must clear the maximum fill height with at least 100 mm of headroom for radar devices, the wetted material must be cross-checked against the process MSDS, and the hazardous-area certificate must match the installed zone [S1][S4].
Shortlist rule of thumb: open atmospheric water/wastewater below 10 m and tight budget, choose hydrostatic or ultrasonic; pressurized chemical, hydrocarbon, or steam service, choose guided-wave radar or non-contact radar; high-accuracy small-vessel custody or additive dosing, choose magnetostrictive; conductive slurries, foam, or low-dielectric media, choose capacitance [S1][S2][S3].
Track two signals through 2026: the IECEx test-report revision cadence for Ex db eb level probes with integrated HART modems, and the rollout of IO-Link wireless level devices for retrofit plants that already run IO-Link masters on valve manifolds, since both will reshape the shortlist for greenfield and brownfield retrofits over the next 12 months [S1][S4].
This topic is covered further in Tariff Relief, Procurement Pitfalls, and a $3.5B Defense Deal: A Cross-Sector Read for....