Guided wave radar (GWR) level meters need a minimum medium dielectric constant around εr ≥ 1.9 to return a usable reflection on standard 316SS cable (6 m) or rod (2 m) probes, as documented for the Signet 2291 TDR/guided wave radar platform [S5]. Below that floor the pulse energy leaks into the vapor space and the unit can no longer resolve level, which is the first hard cut-off an engineer should write into a specification.
For a 2026 selection decision, the dielectric constant of the process medium, not the tank size, is the primary compatibility gate. Coaxial GWR probes handle low-εr liquids down to roughly εr 1.4, twin-rod probes are usable above about εr 1.9, and single-rod probes require a higher εr window, generally 10 or above for clean measurement [S1][S5]. Engineers comparing GWR to non-contact radar level meter options should know the same εr floors apply, although free-space radar pushes the limit lower by focusing the beam.
Dielectric Constant Floors and Probe Geometry
The dielectric constant of the medium sets both the probe geometry and the maximum measurement range of a guided wave radar level installation. For a coaxial GWR probe, the lower dielectric limit is around εr 1.4 because the outer tube contains the electromagnetic field, letting the pulse travel through low-εr hydrocarbons such as LPG, LNG and light oils with low loss [S3]. Twin-rod probes are documented to function down to εr 1.9, and single-rod probes, which radiate the most energy into the open tank, generally need εr of 10 or higher to avoid signal attenuation and false echoes near the probe end [S5].
A common field pitfall is selecting a single-rod probe for a low-εr solvent or LPG service because the datasheet lists εr 1.9 as the published minimum; that 1.9 figure typically refers to a coaxial or twin-rod geometry, not a single rod in free air. The correct step is to match the probe geometry to the medium's dielectric band, then size the probe to within the tank height so it does not bend, touch the bottom, or fall short of the measurement span [S2].
Signal Reliability Versus Free-Space Radar
Free-space 25 GHz K-band radar such as the Signet 2290 covers 0.2 to 18 m with a 19° beam, an SS316Ti horn, and built-in tank mapping, and is rated for fumes, condensation, and light foam [S5]. GWR, by contrast, pulses along a physical probe, so the energy is guided rather than radiated; that gives a stronger return on low-εr liquids but ties the measurement to the mechanical condition of the probe, with build-up, bending, or touching of the tank bottom each becoming a failure source rather than a beam-spread concern [S1][S3].
Compared to a non-contact radar level meter, a GWR TDR level meter typically delivers a more stable reading on liquids whose density or surface turbulence changes, because the pulse travels along a fixed reference path, but it does not perform as well in aggressive foam or heavy coating service, where conductive or coating-resistant build-up can damp the return and shift the apparent level [S1][S3][S4].
Where GWR Wins, and Where It Loses
GWR is the right tool for clean to moderately dirty liquids in the εr 1.9 to 80+ range, including water, acids, caustics, crude, and many solvents, and for short-range interface measurement where the layer above has a substantially different dielectric constant from the layer below [S1][S3]. Its tolerance of liquid density changes and surface turbulence is consistently better than mechanical floats or hydrostatic probes in comparable service, and 1 mm accuracy is achievable in OEM offerings such as the KAMBODA 4-20 mA digital unit with IP68 stainless steel housing [S4].
GWR is the wrong tool where the medium is a non-conductive foam above a low-εr liquid, in tall vessels with heavy internal obstructions, in agitated tanks where a long flexible probe can swing, or in services where heavy coating builds on the probe, because each of those conditions either absorbs the pulse or generates false echoes [S1][S2]. As one engineering source puts it, low-dielectric-constant liquids, foam, and agitated surfaces can each interfere with the return to the point of breaking the measurement [S1].
Installation Pre-Work That Decides Compatibility
Before a guided wave radar is ordered, five site values must be confirmed: tank height, nozzle height, probe clearance, blind zone, and the dielectric constant of the medium at operating temperature and pressure [S2]. Probe length should match tank height: a probe that is too short under-ranges the measurement, and a probe that is too long can contact the tank bottom and generate a permanent false echo at the floor [S2]. For installations where the engineer is laying out cabling and conduit alongside other instrumentation, the single girder crane selection for landfill operations guide illustrates the same "verify envelope dimensions first" discipline applied to a different asset class.
The probe should be mounted in the middle of the tank, away from walls, agitators, heating coils, ladders, support beams, and intake pipes, because any nearby metal object within the pulse path can create false echoes that the GWR electronics will report as level [S2]. The probe should also stay out of the inlet stream, since splashing liquid hitting the probe can produce short-term reflections that read as level jumps; the rule of thumb is to pick a tank location where the surface is relatively calm, even when the rest of the vessel is turbulent [S2]. For dielectric-constant-sensitive services, an on-site dielectric check of the actual process fluid, not the generic datasheet value, is the cheapest insurance against late project surprises.
Output, Materials, and Field Envelope
Standard GWR instruments ship with 4-20 mA analog output, with HART or Foundation Fieldbus options on most OEM platforms, and the loop-powered 2-wire design simplifies retrofit into existing control systems [S4]. Housings are commonly 316SS with IP68 ingress protection, which is sufficient for outdoor mounting and for direct submersion in sumps, and probe materials can be specified as 316SS rod or cable, PTFE-coated cable, or higher alloys for aggressive chemistries [S4][S5].
For comparison, the LibertyCES reference matrix that covers Signet 2280 tuning fork, 2281 conductive multipoint, 2282 guided float, 2284 ultrasonic gap, 2285 float switch, 2290 unguided radar, and 2291 guided wave radar is a useful way to score a service: tuning fork is rated the broadest point-level option because it is independent of conductivity, dielectric constant, viscosity, pressure, and temperature, while the 2281 conductive probe needs at least 10 µS/cm and the 2285 float is positioned for tap water, river water, and sump service [S5]. For continuous level on a low-εr liquid, the 2291 GWR with 316SS cable up to 6 m or rod up to 2 m is the documented pick; for point level on a coating-prone or foaming service, the 2280 fork is the broader match [S5]. Engineers cross-referencing continuous level against acoustic alternatives should also review how the ultrasonic level meter handles foam, since ultrasonic is rated "-" for both dense surface foam and high agitation in the same Signet matrix [S5].
Decision Criteria: Probe Type vs Medium Dielectric
A short selection matrix for continuous liquid level with a guided wave radar probe, where the row is probe geometry and the column is the medium's dielectric band: coaxial probe, εr 1.4 to 10, suitable for LPG, LNG, light hydrocarbons, and most solvents, with the strongest signal but the highest cost and the largest dead band; twin-rod probe, εr 1.9 to 80, the workhorse geometry for clean water, acids, caustics, and many chemicals, with moderate build-up tolerance; single-rod probe, εr 10 and above, the lowest-cost and easiest-to-install geometry for water, slurries, and high-εr process fluids, but the most sensitive to coating and nearby metal [S1][S3][S5].
Two further check values to write into the datasheet review: the published minimum dielectric constant of the chosen probe geometry, and the maximum probe length for that geometry in the medium being measured, since cable probes typically cap at 6 m and rod probes at 2 m for the 316SS standard option [S5]. The trade-off between GWR and non-contact radar in the same service can be framed as follows: a GWR probe returns a larger echo on low-εr liquids but ties the measurement to a mechanical component in the tank, while a free-space 80 GHz radar like the 2290 has no in-tank hardware but requires a clearer beam path and a larger minimum εr to resolve level through vapor [S5]. When in doubt, a side-by-side dielectric measurement at operating temperature, not at 25 °C lab conditions, is the deciding test.
Limits, Failure Modes, and Standards Anchors
The known failure modes of a guided wave radar are: signal loss in low-εr media, false echoes from internal tank obstructions, probe contact with the tank bottom, build-up on the probe shifting the apparent level, and probe swing in agitated service that changes the pulse path [S1][S2]. Mechanical probes also carry a finite risk of breakage from product impact or from improper handling during installation, which is one of the documented failure categories the 3301-series installation guide is built around [S2].
On the standards side, 4-20 mA HART remains the dominant wired signal for continuous GWR transmitters and is FSK-modulated on the analog loop, while Foundation Fieldbus and PROFIBUS PA are fully digital protocols on separate bus segments and are not HART-compatible [S4]. For hazardous-area service, ATEX and IECEx ratings are common on the OEM datasheet, and WHG approval applies to overfill-protection use in some jurisdictions, with the Signet 2280 fork listed at ATEX II 1 G EEx ia IIB T6 as a representative point [S5]. The dielectric-constant check itself is a process-fluid property, not a regulated value, so the published εr floor in the OEM datasheet is the contract between the supplier and the specifier. For dielectric-constant measurement, a separate reference, the specifying a lux meter on an electrical enclosure RFQ guide, is a useful parallel on writing enforceable acceptance criteria into a purchase document.
Track for the next spec cycle: confirmation of the in-service dielectric constant of the actual process fluid at operating temperature and pressure, not at lab ambient, and a documented probe-length vs. tank-height check before any GWR unit is ordered. If the medium sits below εr 1.9, or if heavy coating is expected, switch the evaluation to a non-contact radar level meter or a tuning fork level switch rather than forcing a GWR probe into a service it cannot resolve.