A vortex flow meter has no moving parts and still counts flow. Gas or liquid passes a bluff body, sheds vortices alternately from each side, and the frequency of that shedding is proportional to velocity. A piezoelectric sensor inside the body feels the pressure pulse and turns it into a pulse output. Nothing wears, nothing clogs, and vortex flow meter keeps working on saturated steam at 200 C.
That combination is why this is the default choice for steam in most plants. Orifice plates work on steam too, but they cost permanent pressure loss and they need a long straight run. A vortex flow meter needs 15 diameters upstream and 5 downstream, which is a lot, but it gives back a turndown of 1:10 to 1:15 and a pressure loss you can live with.
On this page
- How the vortex flow meter works
- Specifications
- Straight run and why it is so long
- Steam: the main application
- Gas and compressed air
- Where a vortex flow meter stops working
- Sizing
- Vibration and the sensor that survives it
- Compared with orifice and turbine
- Installation
- Ordering checklist
- Common questions
How the vortex flow meter works
A trapezoidal bluff body sits across the bore. As fluid passes, it separates alternately from the two edges, creating a Karman vortex street downstream. The shedding frequency f relates to velocity v and body width d by a dimensionless number called the Strouhal number, and over a wide Reynolds range that number is essentially constant for a given body shape.
Constant Strouhal number is the whole trick. It means frequency is proportional to velocity with a fixed factor, so one calibration covers the range. Outside that Reynolds range, the number stops being constant and vortex flow meter’s accuracy falls apart, which is exactly what happens at low flow.
The sensor
We use four piezoelectric crystals encapsulated inside the body, behind the bluff body, sensing the alternating pressure. Four crystals instead of one gives a stronger signal and lets the electronics reject common mode vibration. It also means the sensor never touches the fluid, so aggressive condensate and wet steam do not erode it.
The alternative design puts a thermistor or a diaphragm in the flow path. Those are cheaper and they fail faster. On saturated steam, where water hammer is a real event rather than a theoretical one, the encapsulated sensor is the difference between vortex flow meter that lasts ten years and one that lasts two.
Specifications
| Parameter | Range |
|---|---|
| Nominal diameter | DN15 to DN300 (larger on request) |
| Measured media | Steam, gas, liquid (separate calibration per medium) |
| Accuracy | plus or minus 1.0 percent of reading for liquid, 1.5 percent for gas and steam |
| Repeatability | 0.2 percent of reading |
| Turndown | 1:10 standard, 1:15 on selected sizes |
| Medium temperature | -40 C to +400 C |
| Nominal pressure | 1.6 MPa standard, 2.5 / 4.0 MPa optional |
| Body material | SS304 fully welded, SS316 optional |
| Straight run | 15 DN upstream, 5 DN downstream |
| Output | Pulse, 4-20 mA, RS485 Modbus, HART optional |
| Power supply | 24 VDC, or battery for remote steam traps |
| Protection | IP65 standard, IP67 optional |
| Pressure loss | 0.1 to 0.3 bar at full scale, size dependent |
Note the accuracy split. A vortex flow meter is more accurate on liquid than on gas, because gas is compressible and the density term enters the mass calculation. If you need mass flow on steam or gas, add temperature and pressure compensation, which vortex flow meter supports as a built-in option.
Straight run and why it is so long
Fifteen diameters upstream is the number that kills vortex meters in retrofit projects. It exists because vortex flow meter measures a frequency that depends on velocity, and a distorted velocity profile shifts the effective Strouhal number.
Two elbows in the same plane, or a single elbow close upstream, creates a swirl that persists for many diameters. vortex flow meter reads high or low depending on where the sensor sits in the swirl, and the error can reach three to five percent. No amount of electronics fixes it.
Practical options when you do not have 15 diameters:
- Fit a flow conditioner in the available length. A good one recovers most of the accuracy in 5 to 7 diameters.
- Move vortex flow meter further downstream, even if it means a longer spool. Pipe is cheaper than a wrong reading.
- Switch technology. A thermal mass meter needs 1 to 2 diameters and an insertion electromagnetic meter needs 5.
Count the straight run from the flange face, and count only the straight section. A reducer inside the 15 diameters does not count as straight pipe even if it is straight.
Steam: the main application
Saturated steam is where vortex meters earn their place. vortex flow meter handles 400 C, the sensor does not touch the fluid, and the pressure loss is a fraction of what an orifice plate costs you over a year.
Compensation is the part people get wrong. Saturated steam has a fixed pressure-temperature relationship, so you only need one of the two to get density. Most installations measure pressure and compute temperature, which is cheaper and works fine. Superheated steam needs both, because the two are independent.
Watch the dryness fraction. Wet steam carries droplets that erode a bluff body over years, and a vortex flow meter calibrated on dry saturated steam will read high on wet steam because the liquid phase moves faster than the vapour. If your boiler carries water over, fix that before you blame vortex flow meter.
Steam trap monitoring
Battery powered vortex meters with a small bore are used on trap discharge lines to catch failed traps. A trap passing live steam shows continuous flow where there should be intermittent bursts. vortex flow meter pays for itself in a heating season on a large steam main.
Gas and compressed air
On gas the vortex flow meter is a solid choice when the flow stays inside its range and the pipe has room. It handles compressed air, natural gas, nitrogen, oxygen and flue gas, and it does not care about dirt the way a turbine does.
The weakness is the low end. A vortex meter needs a minimum Reynolds number to shed stable vortices, and below that it simply stops counting. On a gas line with a 20:1 day-night swing, the bottom of the swing falls outside the range and you lose it.
If the flow profile has a wide swing, a thermal gas mass flow meter covers 1:100 and reads mass directly. If the gas is clean and the flow is steady, a gas turbine flow meter gives better accuracy at custody transfer points.
Where a vortex flow meter stops working
Low flow is the first limit. Below the minimum Reynolds number there is no stable vortex street, so there is no measurement. vortex flow meter reads zero and looks broken. It is not broken, it is out of range.
High viscosity liquids are the second. A Strouhal number that holds for water does not hold for a 200 cSt oil, because the viscous boundary layer changes the shedding. Above about 30 cSt, use a positive displacement or a Coriolis meter instead.
Very small sizes are the third. Below DN15 the body geometry gets tight, the sensor gets small, and the accuracy claim gets optimistic. For tiny flows, use a thermal mass meter or a small turbine.
And slurry is the fourth. Abrasive particles erode the bluff body edges, which changes the Strouhal number permanently. Once the edges round off, no recalibration brings vortex flow meter back.
Sizing
Size on velocity, not on flow. Each body size has a usable velocity band, and the job is to place the whole operating range inside it with the normal flow near the upper third.
For steam, a useful rule is 15 to 60 m/s. Below 15 the signal gets weak. Above 60 the pressure loss and the erosion rate both climb quickly. For gas, 10 to 70 m/s. For liquid, 0.5 to 7 m/s.
The most common sizing error is picking a body size that fits the pipe rather than one that fits the flow. A DN100 line with a maximum of 15 m3/h of water needs a DN50 meter with reducers, not a DN100 meter that spends its life below the minimum.
Vibration and the sensor that survives it
Vibration is the classic vortex meter problem. A pipe vibrating at 20 Hz looks like a flow signal, and the meter counts it. This is why cheap vortex meters report flow on a shut-off line.
Three defences. The four crystal sensor lets the electronics subtract common mode vibration. The signal processing band-passes around the expected shedding frequency. And the installation should avoid long unsupported spans and compressor discharge lines without pulsation dampers.
If you see a reading on a line that is definitely closed, check the vibration before you check the electronics. Put a hand on the pipe. If you can feel it, the meter can too.
Compared with orifice and turbine
| Vortex | Orifice | Turbine | |
|---|---|---|---|
| Moving parts | None | None | Rotor and bearing |
| Turndown | 1:10 to 1:15 | 1:3 to 1:5 | 1:20 |
| Pressure loss | Low to moderate | High, permanent | Low |
| Straight run | 15 DN up | 10 to 20 DN up | 2 DN up |
| Steam | Excellent | Excellent | Not suitable |
| Dirty fluid | Good | Good | Poor |
| Low flow | Poor | Moderate | Good |
| Cost at DN200 | Moderate | Low | High |
The orifice plate is still the cheapest way to measure steam at large sizes, and it is still the most forgiving of a bad installation. What it costs you is pressure loss, and on a 10 bar steam main that loss is energy you pay for every hour of the year. Run the number before you default to the orifice.
Installation
- Respect the 15 DN upstream and 5 DN downstream. If you cannot, fit a flow conditioner and accept a wider uncertainty.
- Install the meter with the sensor head pointing down on liquid and up on gas, so the electronics stay cool and condensate drains away.
- Insulate the body on steam but leave the head clear. Overheating the electronics is a common cause of early failure.
- Support the pipe on both sides. A vortex meter body is a stress concentration.
- Do not install downstream of a control valve that throttles below 20 percent open. The jet is not a developed profile.
- Ground the body. Static on a gas line upsets the electronics.
Ordering checklist
- Medium: saturated steam, superheated steam, gas or liquid, and the operating pressure
- Maximum, normal and minimum flow with units
- Line size, flange standard and pressure class
- Temperature range, and whether compensation is built in
- Output and protocol, plus whether battery power is needed
- Straight run available upstream and downstream
- Any vibration source within 10 diameters
What the electronics actually do
The sensor gives a weak, noisy signal buried in pipe vibration and electrical interference. Everything between that signal and a 4-20 mA output is filtering, and the quality of that filtering is what separates a vortex flow meter that works from one that reports flow on a closed line.
Three stages matter. A band-pass filter centred on the expected shedding frequency removes most of the vibration energy. An adaptive tracking loop follows the frequency as flow changes, so the filter does not lose the signal during a ramp. And a low-flow cut-off drops the output to zero below a set frequency, which is the only reliable defence against counting vibration as flow.
Set the cut-off honestly. Some installers set it very low so the meter “never misses a reading”, and then the meter reports a phantom flow at night. Set it at the bottom of the calibrated range and accept that the meter reads zero below that, because zero is closer to the truth than a number invented from vibration.
Compensation built in
For steam and gas, the meter can take a pressure transmitter and a temperature sensor directly, compute density, and output mass flow rather than volumetric flow. On saturated steam the pressure input alone is enough, because temperature follows pressure. On superheated steam and on gas, you need both.
Do the compensation in the meter rather than in the DCS. It keeps the flow total correct when the DCS is offline for maintenance, and it means the calibration and the compensation live in the same place when someone audits the numbers three years later.
A worked steam example
A DN80 vortex flow meter on a 10 bar saturated steam main. Maximum flow 3,000 kg/h, normal 1,800 kg/h, minimum 400 kg/h. Is the size right?
At 10 bar saturated, the specific volume is about 0.194 m3/kg, so 3,000 kg/h is 582 m3/h, which through a DN80 bore is roughly 32 m/s. That sits comfortably inside the 15 to 60 m/s band. At the minimum of 400 kg/h the velocity is about 4.3 m/s, which is below the 15 m/s floor.
So the meter covers the maximum and the normal point but drops out at the minimum. Two options. Accept that the meter reads zero below about 1,400 kg/h and use the total for billing, since the total is dominated by the high flow hours. Or fit a DN50 body with reducers, which lifts the minimum velocity to about 11 m/s and still keeps the maximum at 82 m/s, which is too high.
The honest answer in that case is that the range is too wide for one vortex flow meter. Split it into two runs with a small and a large body, or accept the low end loss. That is a decision for the process engineer, not for the meter supplier, and any supplier who says one DN80 covers it has not done the arithmetic.
Sizing table by line size
| Body size | Liquid, m3/h | Gas, Nm3/h | Saturated steam, kg/h at 10 bar |
|---|---|---|---|
| DN25 | 0.4 to 5 | 8 to 60 | 15 to 120 |
| DN50 | 1.5 to 20 | 30 to 250 | 60 to 480 |
| DN80 | 4 to 50 | 80 to 650 | 150 to 1,250 |
| DN100 | 6 to 80 | 120 to 1,000 | 240 to 1,950 |
| DN150 | 14 to 180 | 280 to 2,300 | 540 to 4,400 |
| DN200 | 25 to 320 | 500 to 4,100 | 950 to 7,800 |
Use the table as a first pass and confirm with the actual density. Steam tables and gas composition change the numbers, and vortex flow meter chosen from a table alone is vortex flow meter chosen on an assumption.
Common questions
Why does my vortex meter read zero when the pump is running?
The flow is below the minimum Reynolds number. Check the actual flow against the meter’s range. vortex flow meter sized for the design maximum is often too big for the normal operating point.
Can it measure both steam and water?
Not without recalibration. The Strouhal relationship holds for both, but the density and the compensation differ, and vortex flow meter calibrated for steam will not be accurate on water. Specify the medium at order.
How much pressure loss should I expect?
Roughly 0.1 to 0.3 bar at full scale, depending on size. It is a fraction of what an orifice plate costs, which is the main reason vortex meters replaced orifice runs on steam headers.
Does it work on wet steam?
It works, but it reads high and the bluff body erodes faster. Fix the carryover rather than compensating for it.
Can I install it vertically?
Yes, with upward flow on gas and either direction on liquid, provided the straight run is available. Downward flow on liquid traps air at the sensor.
Vortex flow meter FAQ
How does a vortex flow meter work?
A bluff body sits across the flow. As fluid passes, it sheds vortices alternately from each side, and the frequency of that shedding is proportional to velocity. A sensor detects the pressure pulse from each vortex, and the transmitter converts the frequency into flow rate using the meter’s K-factor.
How much straight run does it need?
15D upstream after a single elbow, up to 30D after a pump or a partly closed valve, and 5D downstream. It is the fussiest of the common technologies on piping layout. Our guide to vortex flow meter straight run requirements has the full table by fitting type.
Can it measure steam?
Yes, and this is where vortex meters earn their place. Saturated and superheated steam both work, provided you specify the right pressure rating and handle temperature and pressure compensation for mass flow. No moving parts means nothing to wear out in hot service.
What is the minimum flow?
It depends on the line size and the fluid density. Below a minimum Reynolds number the vortices stop forming cleanly and the signal drops out. This is why a vortex meter that reads perfectly at full rate can go to zero overnight on a turndown.
Vortex or orifice plate for steam?
Vortex costs more up front but has no impulse lines to block or freeze, and a much lower permanent pressure loss. Orifice is cheaper to buy and easier to calibrate by calculation, but the pressure loss runs for the life of the plant. On a large steam main the energy saving usually pays for the vortex meter within two years.
Does pipe vibration affect it?
Yes. A vortex meter next to a reciprocating compressor will pick up mechanical vibration as signal noise. If the pipe shakes visibly, choose a different technology or isolate the meter section.
Related reading
If your problem is straight run rather than steam, the vortex flow meter straight run guide works through the numbers with drawings. For gas metering with a wide turndown, see the thermal gas mass flow meter. For liquid service on the same line, the liquid turbine flow meter is the alternative when accuracy matters more than dirt tolerance.
Reference for vortex shedding and installation practice: the Karman vortex street is the underlying physics, and the Strouhal number behaviour explains both the accuracy and the low-flow limit.