Eductor Not Mixing? Troubleshoot the System Before You Replace the Part

You have a tank that used to mix and now it doesn’t. Maybe it is louder than it used to be. Maybe the top runs warm while the bottom stays cold, or the pH probe at one end reads differently from the probe at the other. The eductor has no moving parts, so when a mixing loop stops mixing, the instinct is to suspect the eductor. In most cases it is innocent.

An eductor is a shaped nozzle that converts pressure into circulation. Motive liquid enters through a converging nozzle, accelerates, and drops in pressure at the throat, Bernoulli’s equation doing its work. That low-pressure zone drags surrounding tank liquid in through the suction ports, and the diffuser then turns the jet’s velocity back into pressure so the combined stream pushes deep into the tank. The pump supplies the motive stream; the eductor supplies the muscle. There is nothing to break except the flow passages themselves, which is exactly why failures usually turn out to be system problems wearing a part’s face.

This guide is eductor troubleshooting in field order: the five classic failure modes in a symptom-cause-fix table, the installation errors that masquerade as breakdowns, the maintenance checklist that prevents most of them, and the repair-or-replace decision for when the body itself is finally done.

The single most useful habit is to measure at the eductor, not at the pump. Line losses, fittings, valves and an undersized header quietly eat pump head before it reaches the body. A pump delivering 5 bar at its flange can present 3 bar or less at an eductor twenty meters of pipe away. Most eductor troubleshooting calls are answered by three readings: the pressure at the eductor inlet, the liquid level above the suction ports, and the condition of the throat.

The second habit is a baseline. You cannot troubleshoot a loop you never measured. At commissioning, log the motive pressure at the body, the turnover time (how long it takes to circulate one tank volume), and the sound of the discharge. Every later symptom is a deviation from that baseline: a turnover time that doubled is a diagnosis even when nothing looks wrong.

The Five Classic Failure Modes

Start with the map, then read the section that matches your symptom. These five failure modes account for nearly every eductor service call: insufficient circulation, cavitation noise, low discharge pressure, clogging, and wear.

Symptom	Likely cause	Check first	Fix
Weak or dead circulation	Motive pressure below the body’s minimum	Pressure at the eductor inlet, not the pump	Raise pump output, shorten the line, or re-size
Rattling, gravel or gurgling noise	Cavitation or air entrainment	Liquid temperature vs vapor pressure; submergence	Restore submergence; bring pressure back into the window
Discharge can’t reach across the tank	Worn throat, misaligned nozzle, or back-pressure	Throat bore, nozzle seating, tank level	Replace wear parts; re-aim; re-size for the head
Flow fell off suddenly	Clogged throat or suction ports	Strainer, throat, ports	Clean; fix the strainer; manage solids
Flow up, mixing down	Throat and nozzle erosion enlarging the passages	Dimensional check vs as-new baseline	Replace wear parts; upgrade material

Read the last row twice: with an eroded eductor the flow often rises while the mixing falls, because a bigger throat at the same pressure means lower jet velocity, and entrainment runs on jet momentum. If your flow meter says more and your tank says less, you are looking at wear, not clogging.

Weak Circulation: The Pressure Story

The motive pressure at the body sets everything. Below the design minimum the jet is weak, the suction zone is shallow, and the induction ratio collapses. The eductor keeps passing liquid, but it stops entraining the tank. This is the most common “the eductor stopped working” report, and it is almost always a pressure problem upstream of the body.

Work through the pressure chain in order. Is the pump healthy and running on its curve? Is the strainer on the motive line clean. A clogged strainer is a pressure loss hiding as a pump problem? Are the isolation valves fully open? Is the header sized for the total motive flow, or is one undersized line feeding three eductors and starving them all? Is the run to the tank long enough that friction eats the head? Each of these quietly lowers the pressure that actually reaches the eductor.

The math explains why pressure matters more than flow here. Through a fixed nozzle, flow scales with the square root of pressure: doubling the pressure raises the flow only about 41%. But entrainment depends on jet momentum, which scales with the square of velocity. A 20% loss of pressure at the body is a modest flow loss and a large mixing loss, which is why a loop can pump plenty of liquid and still fail to mix.

The induction ratio itself is a range, not a spec point. It falls as the discharge fights back-pressure, the liquid head above the body in a tall tank, and it falls as the liquid thickens, because a more viscous fluid resists being dragged into the jet. Size for the worst case you will actually see: highest tank level, most viscous batch. A body sized for the best day is a body that quits on the worst one.

Two air problems masquerade as pressure problems. If the pump is sucking air, a leak on its suction side, a low sump level, the motive stream carries gas, and gas absorbs momentum instead of transferring it; the suction at the ports dies even with the gauge reading fine. And an eductor in a multi-body header can die alone from an air lock or a partly closed valve while its neighbors run normally. Purge and vent the header, walk the valves, and a “failed” eductor is often back to work in minutes.

A quick sizing sanity check for weak circulation: a 20 m³ tank at three turnovers per hour needs about 60 m³/h of circulated flow. At a 5:1 induction ratio that means roughly 12 m³/h of motive flow. If your pump delivers only half that at the body, no amount of cleaning will fix the mixing, and the answer is a higher induction ratio (more pressure) or more eductors on the header.

Cavitation Noise: The Throat Is Singing

An eductor is never silent. The steady hiss of a coherent high-velocity jet is the sound of the device working. The throat is doing exactly what it was designed to do. What you should not hear is rattling, a gravel sound, or a wet gurgle. Those are cavitation or air, and both damage the body and collapse the mixing.

Cavitation is the local boiling of the liquid inside the body. The throat is the point of maximum velocity and minimum static pressure; if that pressure falls below the liquid’s vapor pressure, vapor bubbles form in the stream and collapse violently as the pressure recovers in the diffuser. The collapse is what you hear, and it is what erodes metal. Cavitation pitting looks like a sandblasted, frosty surface, distinct from the smooth polish of abrasive wear.

Three conditions push a body into cavitation. The first is the liquid itself: hot water and volatile solvents sit close to their vapor pressure, so the margin between throat pressure and boiling is thin. The second is excessive motive pressure: above the body’s design window, the extra velocity drops the throat pressure further, and noise and pitting follow. The third is air entrainment: a vortex at the suction ports pulling gas into the stream changes the whole flow regime and produces the same rattle.

Diagnose in that order. Check the liquid temperature against its vapor pressure at operating conditions. A hot batch is the classic trigger for an eductor that cavitates only on certain days. Measure the motive pressure at the body and compare it to the vendor’s window: if it is above the window, throttle it back rather than assuming more pressure is better. Then check submergence, because the cheapest cavitation fix is stopping the vortex that should never have started.

Fix the cause before the parts. Restore submergence, throttle the pressure into the window, and if the duty genuinely runs hot, re-check the body selection. A hot service needs a body sized for the actual vapor-pressure margin, not the cold-water assumption. Replace pitted throats and nozzles afterward; a pitted surface erodes far faster than a smooth one.

Low Discharge Pressure: The Diffuser’s Job

The discharge stream is supposed to carry momentum to the far end of the tank. That is what makes an eductor a mixer instead of just a pump. When the stream dies a meter out of the body, or when an eductor used for jet pumping loses its lift, the diffuser is not doing its job.

The diffuser converts the jet’s velocity back into pressure. Its geometry is critical: a typical liquid eductor uses a diffuser half-angle around 5–7°, steep enough to recover pressure quickly, shallow enough to avoid flow separation. If separation occurs, from an oversized throat, a worn profile, or gas in the stream, the pressure recovery collapses and the discharge dribbles instead of pushing.

The usual suspects, in order: worn throat and diffuser (erosion opens the passages, jet velocity falls, recovery falls with it); a nozzle that was not reseated properly after maintenance (the alignment of nozzle, throat and diffuser is what makes an eductor work, and reassembly marks exist for a reason); motive pressure below the window; gas or air in the stream; and a discharge head the body was never sized for. A body selected for a half-full tank cannot overcome a full one.

Check the aim before you condemn the hardware. A discharge aimed upward churns the surface and leaves the floor dead; one aimed at the near wall creates a dead zone behind the body; the same body aimed along the floor across the long axis will sweep the whole volume. Re-aiming costs nothing and fixes a surprising share of “weak discharge” reports.

If the eductor is on jet-pump duty, lifting liquid from a sump or pit, typically three to six meters, the symptom shows as lost lift. The causes are identical: worn internals, low motive pressure, or gas in the motive stream. Measure the pressure at the body, inspect the wear parts, and only then re-spec.

Clogging: Solids Are a System Problem

The throat is the smallest passage in the loop, which makes it the natural collection point for anything that should not be in the line. A clogged eductor fails suddenly, flow falls off, suction at the ports dies, and often the sound changes, which is your first clue to the diagnosis. Clogs are abrupt; wear is a drift. The timeline alone separates the two.

Trace the solids. Is there a strainer on the motive line, and is it intact and the right mesh? A torn strainer or a coarse one lets the debris through that the throat then catches. Are the suction ports pulling settled solids off the tank floor. A body mounted down in the sludge bed will draw sludge no matter how clean the motive line is? Is the liquid prone to scaling or crusting that builds up in the throat over weeks rather than jamming it overnight?

Clearing a clog is usually straightforward. Soft blockages often clear with a reverse flush, run the discharge line back through the body, but check the vendor’s guidance first and mind any check valve that would make reverse flow impossible or dangerous. Hard blockages need the body out and the throat cleaned mechanically. The more important step is the fix behind the fix: repair or replace the strainer, mount the body above the settled solids bed and let the discharge sweep it, and if solids keep recurring, the honest answer is that the body is too small for the duty and a larger open-passage design is the cure.

Wear: Flow Goes Up, Mixing Goes Down

Erosion is the only wear an eductor knows, and it happens in exactly the places you would predict: the nozzle, where velocity is highest; the throat, where the jet is concentrated; and the entrance of the diffuser, where the stream turns. Abrasive solids accelerate it, high velocity multiplies it, erosive wear scales with velocity squared, and cavitation pitting adds a second, faster damage mechanism on top.

The wear signature is counterintuitive, which is why it gets misdiagnosed. As the throat and nozzle erode, the passages get larger. At the same motive pressure, a larger passage passes more flow, but the jet velocity falls, and entrainment runs on jet momentum. So the loop’s flow meter climbs while the mixing quality drops. If someone reports “more flow, worse mixing,” the part is not clogged; it is worn out.

Check wear dimensionally, not by feel. Measure the throat bore and nozzle diameter against the as-new figures from the vendor and log the trend; a body that has drifted a few percent past its tolerance is costing you mixing on every batch. Inspect the surfaces for the frosty pits of cavitation versus the smooth enlargement of erosion. They point to different causes and different fixes.

Replace the wear parts when they drift past tolerance, and do it on an inspection schedule rather than on failure. A worn throat rarely announces itself before a batch is ruined. If the duty is abrasive, upgrade the material at the same time: hardened alloys and ceramic throats cost more up front and pay for themselves the first season they survive. And treat the cause: settle or filter the abrasive load, keep the suction ports out of the sludge, keep the pressure inside the window. Replacing a throat every month is not maintenance; it is a symptom you have not diagnosed yet.

Installation Errors That Look Like Breakdowns

A large share of “broken” eductors are simply installed wrong, and the symptoms are indistinguishable from genuine failures. Orientation, depth and spacing are the three that dominate.

Installation error	What you see	The fix
Discharge aimed upward	Surface churns, floor stays dead, “no mixing”	Re-aim along the floor or slightly downward
Discharge aimed at the near wall	Dead zone behind the body	Aim across the long axis to sweep the volume
Body mounted above the operating low level	Vortex, gurgling, air in the suction	Mount below the worst-case low level, not the best case
Bodies spaced too close	Streams interfere and cancel each other	One eductor per 2–3 m of tank length, discharge streams reinforcing
Bodies spaced too far apart	Dead zones between them	Add bodies or re-aim to overlap the sweep
Motive fed to the wrong port, valve partly closed	One eductor dead in a header while others run	Walk the valves, purge and vent the header
Nozzle not seated after maintenance	Weak discharge, poor induction	Reseat per the vendor’s assembly marks

The classic field story is the loop that “works on Monday and gurgles on Friday”: the body was mounted to clear the liquid at the normal level, and when the tank drains toward the weekend, the body breaks the surface and starts pulling a vortex. The fix is to mount for the worst-case low level, not the comfortable one. Level swings are the most common silent cause of intermittent eductor failures.

The Maintenance Checklist That Prevents Most Calls

The eductor’s maintenance schedule is measured in years, not months, but it is not zero. A quarterly walk-down catches nearly every failure before it costs a batch.

Repair or Replace: When the Body Is Done

An eductor body is, in effect, a carrier for wear parts. If the only damage is the throat or nozzle and the body is sound, replace the parts. That is what they are for. The decision turns on the condition of the body itself and the duty it now faces.

Condition	Repair	Replace
Throat or nozzle eroded past tolerance, body sound	Replace wear parts	-
Body cracked, corroded through, or leaking	-	Yes
Repeated clogging despite a correct strainer	-	Yes: re-spec for a larger open passage
Duty changed: viscosity, flow, or discharge head moved	-	Yes: re-size the body for the new numbers
Repair labor approaching the cost of a new body	-	Yes
Duty is abrasive or hot and the current material keeps failing	-	Yes: with an upgraded material

The rule that saves the most money: never replace the part before you have ruled out the system. The most common “the new eductor still doesn’t work” story is an old system still attached to it: same low pressure at the body, same under-submergence, same clogged strainer, new part. Diagnose the loop, then buy the part. And when the duty itself has changed, the tank got taller, the liquid got thicker, the turnover requirement doubled, replacing the eductor with the same model is not a repair, it is a repeat. Re-size from the current duty before you order anything.

A Field Diagnostic Sequence

When the loop misbehaves, run the checks in this order and stop at the first hit:

Confirm the symptom against the baseline: turnover time, sound, pressure trend. A doubled turnover time is already a diagnosis.
Read the pressure at the eductor inlet. Below the window, chase the system first: strainer, valves, pump, line losses.
Check the liquid level against the body position. Watch for a vortex at the suction ports.
Check the strainer and the throat for blockage. A sudden failure points here; a drift points at wear.
Walk the header, valves, venting, air locks, especially on multi-body loops where one eductor is dead alone.
Inspect the wear parts dimensionally against the as-new figures.
Only then decide repair versus replace, using the duty, not the catalogue, as the tiebreaker.

Frequently Asked Questions

Why is my eductor so loud? A steady hiss is normal. It is the high-velocity jet doing its job. Rattling or a gravel sound is cavitation or air: check liquid temperature against vapor pressure, motive pressure against the window, and submergence. A gurgle is usually a vortex at the suction ports.

Why did my tank stop mixing overnight? The three classic overnight killers: the level dropped and the body broke the surface (vortex), the strainer plugged and the pressure at the body collapsed, or someone closed a valve. The pressure-at-the-body reading and a look at the level answer all three in minutes.

Why does my eductor work on Monday and gurgle on Friday? That is a level-swing story. The body is mounted above the worst-case low level, so as the tank drains it starts pulling air. Re-mount below the low level. This is the single most common intermittent eductor failure.

Why is my flow higher but my mixing worse? That is the wear signature: an eroded throat and nozzle pass more flow at the same pressure, but the jet velocity falls and entrainment collapses. Measure the throat dimensionally; the part is worn, not clogged.

Can I unclog an eductor by reversing the flow? Soft blockages often clear with a reverse flush, but check the vendor’s guidance first and respect any check valve. Hard blockages need the body out and the throat cleaned. Then fix the strainer. The clog was only the messenger.

How often should I replace the eductor throat and nozzle? On an inspection schedule, not on failure. Replace when the dimensions drift past the vendor’s tolerance. Abrasive duty shortens the interval; that is a signal to upgrade the material or clean up the solids, not to accept monthly replacements.

When should I just buy a new eductor? When the body itself is damaged (cracked, corroded, leaking), when the duty has changed and the old size no longer fits, or when repair cost approaches replacement cost. And never before you have ruled out the system. A new eductor on the same bad system is the same bad loop.

If the loop is still weak after working through the checks, or if the duty has changed and the body needs re-sizing, send the tank geometry, the liquid properties, and the pump curve to the engineering desk and the sizing will be worked from your numbers, not the catalogue. The eductor nozzle range covers tank-mixing and jet-pump bodies, the what is an eductor guide walks through the principle and the sizing arithmetic, and the small-pump tank mixing guide covers running a full loop on a pump you already own.