Most flow meter selection goes wrong at the first step. Someone asks for flow meter selection guide for a DN100 line, and the conversation starts with pipe size when it should start with the fluid. Pipe size is the last thing that matters and it is the first thing people mention.
This flow meter selection guide works in the order an engineer actually decides: fluid first, then range, then accuracy, then pipework, then budget. Each step eliminates options, and by the fifth step there are usually only two left.
On this page
Step 1: the fluid
The fluid eliminates more options than any other factor. Gas, steam, clean liquid, dirty liquid, slurry and viscous product each have a short list, and the lists barely overlap.
Gas rules out every positive displacement meter except the rotary type, and it rules out electromagnetic and Coriolis on cost. Steam rules out anything with a rotor or a sensor in the flow. Clean liquid keeps almost everything. Slurry keeps only electromagnetic, Doppler ultrasonic and a few specialised designs. Viscous product keeps positive displacement and Coriolis.
Ask three questions about the fluid. Does it conduct electricity? Does it carry solids? Does it change phase anywhere in the process? Those three answers cut the field by half.
Step 2: the range
The turndown you need is the ratio of maximum to minimum flow, and it is the second biggest eliminator. A line that swings from 5 to 100 cubic metres per hour needs a 1:20 turndown. An orifice plate cannot do it. A turbine can. An electromagnetic flowmeter can do 1:100.
Get the minimum right. The minimum is not the design minimum, it is the lowest flow that actually occurs, including night-time, weekends and shutdown purge. flow meter selection guide sized on the design minimum loses the quiet hours, and the quiet hours are often a third of the total volume.
Also check the direction. If reverse flow is normal rather than an upset, say so, because a unidirectional meter will read zero or be damaged.
Step 3: the accuracy requirement
Accuracy is where money gets spent for no reason. A custody transfer point needs half a percent and a calibrated certificate. A cooling water branch that feeds a reporting dashboard needs two percent and nothing more.
Ask what the number is used for. Billing, invoicing, regulatory reporting and allocation between parties need accuracy and traceability. Trend monitoring, leak detection and energy reporting usually do not.
The gap between half a percent and one percent is often a factor of two in price, and it is frequently bought by people who do not need it. That money is better spent on a better installation, because installation error dwarfs instrument error in most plants.
Step 4: the pipework
Now, and only now, look at the pipe. Straight run available, pipe size, whether it can be shut down, and whether it can be cut.
| Constraint | What it eliminates |
|---|---|
| Under 5 D upstream | Vortex, orifice, most differential pressure |
| Cannot shut down | Everything except clamp on ultrasonic |
| Above DN600 | Full bore electromagnetic on cost |
| Below DN15 | Vortex, ultrasonic, orifice on accuracy |
| Vertical downflow on liquid | Most meters, needs relocation |
| Buried pipe | Anything needing access, unless IP68 and remote electronics |
The straight run constraint is the one that surprises people, because it is invisible on a P and ID. A vortex meter needs fifteen diameters and a turbine needs two, and the drawing looks the same either way.
Step 5: the budget
Budget is last because by this point there are two or three candidates, and the comparison is meaningful. Compare total installed cost, not purchase price. A cheap meter that needs a spool piece fabricated and a shutdown is not cheap.
Include the transmitter, the cabling, the isolation valves, the bypass and the engineering time. On a small line flow meter selection guide is often a quarter of the installed cost, and on a large line it is most of it.
Then add the running cost. A differential pressure meter on a steam main costs pressure every hour of the year. flow meter selection guide with a rotor costs a bearing every few years. flow meter selection guide with no moving parts costs a calibration every five years.
Technology comparison
| Meter | Best for | Turndown | Accuracy | Weakness |
|---|---|---|---|---|
| Electromagnetic | Conductive liquid, slurry | 1:100 | 0.5 percent | Needs conductivity |
| Ultrasonic clamp on | Retrofit, large pipe | 1:50 | 1.0 percent | Pipe data critical |
| Turbine | Clean liquid, custody | 1:20 | 0.5 percent | Dirt and bearings |
| Vortex | Steam and gas | 1:15 | 1.0 percent | Straight run, low flow |
| Thermal mass | Gas and compressed air | 1:100 | 1.0 percent | Not for steam or liquid |
| Orifice | Steam, high pressure | 1:5 | 0.5 percent | Pressure loss |
| Balanced orifice | Steam, short run | 1:8 | 0.5 percent | Cost, turndown |
| Positive displacement | Viscous liquid | 1:100 | 0.5 percent | Cost, dirt |
| Coriolis | Mass, viscous, small flow | 1:100 | 0.2 percent | Cost, size limit |
Read the table by eliminating, not by ranking. The right meter is the one whose weakness does not matter for your duty, and every meter has a weakness.
Three worked examples
A 10 bar saturated steam main, DN150, 8 tonnes per hour maximum, 1.2 tonnes minimum. Fluid is steam, so turbines are out. Turndown is 1:6.7, so an orifice is marginal and a vortex is comfortable. Straight run is 8 diameters, so a standard vortex is out and a balanced orifice fits. Accuracy requirement is allocation, not billing. Answer: balanced orifice, or a vortex with a flow conditioner if the run can be extended.
A diesel loading line, DN80, 40 cubic metres per hour maximum, 8 minimum. Fluid does not conduct, so electromagnetic is out. Turndown is 1:5. Clean liquid, so turbine is in. Answer: liquid turbine flow meter, 0.5 class, with a 100 micron filter.
A cooling water header, DN300, 200 cubic metres per hour, no shutdown available. Fluid conducts, but the pipe cannot be opened. Answer: clamp on ultrasonic flow meter, two path, with measured wall thickness. The electromagnetic flowmeter would be more accurate and cannot be installed.
Each example took four questions and produced one answer. That is what a flow meter selection guide is for, and none of them started with pipe size.
Checking the answer against a real installation
Before the purchase order goes out, find flow meter selection guide doing the same job somewhere in the plant and look at it. This is the single most useful step in any flow meter selection guide and almost nobody does it.
Look at three things. Where it is mounted, how much straight run it actually has, and whether the operators trust the reading. flow meter selection guide that has been in service for five years tells you more about the technology than any datasheet.
Ask the maintenance team what fails. flow meter selection guide that needs a bearing every two years is a different proposition from one that needs nothing, and the difference does not appear in the price.
If there is no reference installation, ask the supplier for one. A supplier who cannot name a plant running the same meter on the same duty is a supplier who is guessing.
When the fluid changes
Flow meter selection is usually done for the process as designed. Processes change, and flow meter selection guide that was right at commissioning is sometimes wrong five years later.
A new product with a different viscosity changes the turbine requirement. A change from water to a solvent removes conductivity and rules out the electromagnetic flowmeter. A process change that lowers the minimum flow pushes a vortex meter below its working range.
Revisit flow meter selection guide selection guide whenever the process changes, not just when flow meter selection guide fails. The failure mode of a mismatched meter is a slow drift that nobody notices, which is worse than a hard failure.
The other change worth watching is the operating pressure. flow meter selection guide selected at 4 bar and now running at 1.5 bar may be outside its range on gas, because the actual volume flow has risen by a factor of nearly three.
How to compare two quotations
Two quotations for the same duty rarely describe the same scope. Compare them line by line against this list.
- Is the transmitter included, and is it the range you asked for
- Are isolation valves and a bypass included
- Is the straight run spool included, or assumed to be site fabricated
- Is the calibration certificate traceable, and to what standard
- Are flanges to the plant standard or to the supplier’s standard
- Is installation supervision included
- What is the lead time, and what is the spares lead time
- What is the warranty, and does it cover the liner or the bearing
The cheapest quotation is usually the one with the smallest scope, and the difference shows up as site work. Add the missing items to both and compare again, and the ranking often reverses.
Five selection mistakes
Starting with pipe size. It is the last variable, not the first, and starting there locks in a bad answer.
Sizing on design maximum and ignoring the minimum. flow meter selection guide that covers the peak and loses the night is flow meter selection guide that measures two thirds of the volume.
Buying accuracy you do not need. Half a percent on a trend display is money spent on nothing.
Ignoring installation cost. The installed cost is what the budget approves, not the purchase price.
Forgetting the fluid properties. Conductivity, viscosity, solids and phase changes decide more than any datasheet figure.
What to send an enquiry
- Fluid, concentration, temperature, pressure and conductivity
- Maximum, normal and minimum flow, with units
- Whether the number is for billing, allocation or indication
- Line size, material and flange standard
- Straight run available upstream and downstream
- Whether the line can be shut down and cut
- Output and protocol required
- Hazardous area classification
Send those eight items and any supplier can quote accurately. Send three and you will get three different meters, and the cheapest one will be wrong.
Where this flow meter selection guide stops
This flow meter selection guide covers the common industrial cases. It does not cover three situations where the answer is different and the stakes are higher.
The first is custody transfer. Any point where money changes hands on the reading usually falls under a local metrology rule that specifies the permitted technology, the calibration interval and the certificate format. No flow meter selection guide can override that, and flow meter selection guide that is technically better may still be non-compliant.
The second is safety instrumented service. A flow meter used in a safety function needs a declared failure mode, a proof test interval and a certified safety integrity level. That is a different product from the process meter beside it, even when the tag looks the same.
The third is two-phase and multiphase flow. Wet gas, steam with condensate and oil-water-gas mixtures need dedicated solutions, and a flow meter selection guide that pretends otherwise is dangerous. If the fluid changes phase anywhere in the measurement section, stop and get specialist advice.
Reviewing a selection before you buy
Run this review before the order goes out. It takes twenty minutes and it catches the mistakes that cost the most.
Check the minimum flow against flow meter selection guide’s range, not against the design minimum. Check the turndown against the real swing, including shutdowns. Check the fluid properties at the operating temperature, not at ambient. Check the straight run on the installation drawing, not on the P and ID. Check that the accuracy class matches what the number is used for. Check that the output protocol matches what the control system can accept. Check the lead time against the shutdown date.
Then check the whole thing against a reference installation one more time. A flow meter selection guide is a checklist, and a checklist only works when someone actually runs it.
What the budget actually covers
A flow meter selection guide that stops at the purchase price is only half useful. The installed cost is what gets approved, and it is built from six items that are easy to forget.
flow meter selection guide itself. The transmitter, if it is separate. The spool piece, if flow meter selection guide needs one, and the flanges to match the plant standard. The isolation valves and the bypass, which turn a meter change into a two hour job instead of a shutdown. The cable, the gland and the control system input card. And the engineering time to produce the installation drawing and update the as-built.
On a small line flow meter selection guide is often a quarter of the installed cost. On a large line it is most of it. Either way, the comparison only means something when both quotations are costed on the same scope.
Add the operating cost too. A differential pressure device on a steam main loses pressure every hour of the year, and that loss is energy with a price. A meter with a rotor needs a bearing every few years. A meter with no moving parts needs a calibration every five years and nothing else.
Accuracy class and what it costs you
Every step up in accuracy class roughly doubles the price, and the money is usually better spent elsewhere. This is the part of a flow meter selection guide that saves the most budget.
Two percent is enough for trend monitoring, leak detection and most energy reporting. One percent covers allocation between departments and most internal charging. Half a percent is for custody transfer, invoicing and regulatory reporting. Anything tighter than half a percent is Coriolis territory and only worth it where the product value justifies it.
Compare that with installation error. A meter with a half percent certificate installed with two diameters of straight run and a guessed wall thickness reads three percent wrong. A meter with a one percent certificate installed properly reads one percent wrong. The installation decides the outcome, not the certificate.
So the question to ask is not what accuracy flow meter selection guide has, but what accuracy the installation will deliver. On a line with generous straight run and a clean fluid, buy the better meter. On a tight retrofit, buy the cheaper meter and spend the difference on getting the installation right.
Common questions
What is the most common flow meter?
By unit count, the electromagnetic flowmeter in water and wastewater, and the orifice plate in steam and process. By application variety, the electromagnetic flowmeter covers the widest range of fluids.
Which flow meter is most accurate?
Coriolis, at around 0.1 to 0.2 percent, and it measures mass directly. It is also the most expensive and it is limited in size, which is why it is used on small and medium lines where accuracy matters most.
Which flow meter needs the least straight run?
A turbine, at two diameters, because the integral straightener does the work. A precession vortex meter needs three, a thermal mass meter needs one to two, and a standard vortex needs fifteen.
Can one meter measure both gas and liquid?
Not practically. Coriolis can, and it costs accordingly. Everything else is specified for one phase, and using it on the other gives a wrong reading or a damaged meter.
How do I choose between a turbine and an electromagnetic flowmeter?
Conductivity decides it. If the liquid conducts, the electromagnetic flowmeter handles dirt and slurry and has no wear parts. If it does not conduct, the turbine is the economical choice for clean product.
What flow meter works with no straight run?
None, strictly. A clamp on ultrasonic meter is the most forgiving because it does not disturb the flow, but it still needs about ten diameters for accuracy. A precession vortex meter at three diameters is the shortest practical requirement.
Related reading
For flow meter selection guides referenced here, see the electromagnetic flow meter, the vortex flow meter, the thermal gas mass flow meter and the balanced orifice flow meter. If your meter is already installed and reading wrong, start with flow meter troubleshooting.
Reference for selection and installation practice: ISO 5167 covers differential pressure devices and the straight run rules that most other technologies are compared against.
Working out velocity from your flow and bore? Use the flow rate and pipe velocity calculator before you settle on a line size.
Comparing technologies side by side? The flow meter technology comparison sets 11 types against 12 selection criteria in one table.