Measuring level with differential pressure and a wet reference leg: correctly calculate wet leg, density and zero shift

Differenzdrucktransmitter an geschlossenem Behälter mit gefüllter Wet Leg Referenzleitung zur Füllstandmessung
→ Product category: Level measurement

The level in a closed process vessel is to be measured using a differential pressure transmitter. The high-pressure side is connected to the lower vessel connection, while the low-pressure side is connected to the gas space. This means that the gas-space pressure acts on both sides of the transmitter and can largely cancel out when the differential pressure is calculated.

In many applications, however, the reference line to the upper vessel connection does not remain dry. With steam, hot media or condensable gases, liquid can form in the line. A partially filled line would be particularly unfavourable for the measurement because the additional hydrostatic liquid column would then change uncontrollably. Instead of leaving this condition to chance, the reference line is therefore deliberately filled completely with a defined liquid. This is referred to as a wet reference leg or wet leg.

This makes the measurement reproducible – but it changes the calculation. The liquid column in the wet leg acts permanently on the low-pressure side of the differential pressure transmitter. Even with an empty tank, a significant negative differential pressure can therefore be present. This value is not an instrument fault, but a deliberately created offset of the measuring point.

The density of the reference liquid is also decisive. If its temperature changes, its density changes as well, and therefore so does the hydrostatic counterpressure. If the reference line is not completely filled, gas bubbles are present or the liquid level in the condensate pot drops, the indication will also shift.

The most important rule is therefore: In wet-leg level measurement, the process liquid and reference column must be balanced separately. The height and density of the wet reference leg determine the zero shift, while the density and level height of the process medium determine the actual measuring span.

How does level measurement with differential pressure work?

The hydrostatic pressure of a liquid column is determined by its height and density. In simplified form:

p = ρ × g × h

Where:

  • p = hydrostatic pressure,
  • ρ = density of the liquid,
  • g = gravitational acceleration,
  • h = height of the liquid column.

In an open vessel, a pressure transmitter at the lower measuring point can measure this hydrostatic pressure directly. In a closed vessel, however, an additional gas or vapour pressure acts on the liquid surface.

The pressure at the lower vessel connection is then approximately:

pbottom = pgas + ρP × g × h

A single gauge pressure transmitter could not distinguish which portion of this pressure is caused by the liquid level and which by the gas space.

A differential pressure transmitter is therefore used. The high-pressure side H or + is connected to the lower process connection. The low-pressure side L or − receives the gas-space pressure from the upper connection.

With an ideally dry reference leg:

Δp = (pgas + ρP × g × h) - pgas

and therefore:

Δp = ρP × g × h

The gas-space pressure cancels out. This same basic principle also applies with a wet leg. However, an additional liquid column acts on the low-pressure side.

What is a wet reference leg or wet leg?

A wet leg is a deliberately liquid-filled reference line between the gas space of a closed vessel and the low-pressure side of a differential pressure transmitter.

The liquid may, for example, consist of condensate from the process medium. In other applications, a separate suitable sealing or fill liquid may be used.

The decisive factor is not simply that some liquid is present in the line. For reproducible measurement, the reference column must have a defined and as constant a condition as possible.

A half-filled line is significantly worse from a measurement perspective than a deliberately dry or deliberately completely filled line. If the liquid level in the reference line moves with the operating condition, the additional hydrostatic pressure on the L side changes and directly affects the indicated level.

A wet leg is therefore used particularly where condensation is expected anyway. Instead of allowing uncontrolled condensate formation, a fixed reference column is created.

Distinguishing between dry leg and wet leg

Characteristic Dry Leg Wet Leg
Condition of the LP reference line Gas-filled or dry Completely filled with liquid
Typical application Non-condensing gas or dry gas space Condensing vapour or process gas with expected liquid formation
Additional hydrostatic pressure on LP side Approximately none Constant liquid column
Differential pressure at 0% Often approximately 0 with ideal geometry Often significantly negative
Important additional parameter Reference line must remain free of condensate Wet-leg height and density of the reference liquid
Typical source of error Unintended condensate formation Partial draining, gas bubbles, density change or changing reference temperature

The key point is reproducibility. A dry-leg arrangement works as calculated only if the line actually remains dry. A wet leg works as calculated only if the reference column is actually complete and has the assumed density.

Basic equation for wet-leg measurement

For a simplified arrangement, it is initially assumed that the differential pressure transmitter is installed at the same height as the lower process connection. Additional height effects of the high-pressure line are initially neglected.

On the high-pressure side:

pH = pgas + ρP × g × h

On the low-pressure side:

pL = pgas + ρW × g × HW

This gives:

Δp = pH - pL

and therefore:

Δp = ρP × g × h - ρW × g × HW

Where:

  • ρP = density of the process medium,
  • h = current process level above the lower measuring point,
  • ρW = density of the wet-leg liquid,
  • HW = vertical height of the wet-leg liquid column.

The gas-space pressure appears in both equations and cancels out when the differential pressure is calculated.

The wet-leg column, however, remains. Because it acts on the low-pressure side, its hydrostatic pressure is subtracted from the useful signal.

Why 0% level can produce a negative differential pressure

With an empty vessel, or at 0% of the defined measuring range, the hydrostatic process column in the simplified example is zero:

h = 0

This leaves:

Δp0% = -ρW × g × HW

The differential pressure transmitter therefore sees a negative value because the filled L side produces a higher hydrostatic pressure than the H side.

This is completely normal.

An empty tank in a wet-leg arrangement therefore does not have to generate 0 mbar differential pressure. A level of 0% may, for example, correspond to −400 mbar, −300 mbar or another project-specific value.

This constant offset is taken into account during parameterisation. In classic transmitter terminology, such a shift is often referred to as zero elevation.

The terminology itself is less important than the correct calculation: the physically present differential pressure must not simply be “zeroed out” by an incorrect zero adjustment without also taking the corresponding range limits into account.

Correctly calculating LRV and URV

For parameterising a modern differential pressure transmitter, the Lower Range Value and Upper Range Value are particularly important.

With the simplified geometry, at 0%:

LRV = -ρW × g × HW

At 100% level with measuring height HP:

URV = ρP × g × HP - ρW × g × HW

The actual measuring span is:

Span = URV - LRV

This gives:

Span = ρP × g × HP

This is an important relationship: an ideally constant wet-leg column shifts the entire differential pressure range but does not change the span generated by the process level.

This is exactly why, in a wet-leg arrangement, both LRV and URV can be negative while the transmitter still outputs a perfectly normal 4–20 mA level signal from 0 to 100%.

Practical example with a 4 m wet leg

A closed process vessel is to be measured over a level range of 3.0 m. For this simplified example, the differential pressure transmitter is installed at the same height as the lower measuring connection.

Given:

  • measuring height of the process medium: 3.0 m,
  • density of the process medium: 800 kg/m³,
  • height of the wet reference leg: 4.0 m,
  • density of the wet-leg liquid: 1,000 kg/m³,
  • gravitational acceleration: 9.80665 m/s².

The hydrostatic pressure of the completely filled reference line is:

pWetLeg = 1,000 × 9.80665 × 4.0

pWetLeg ≈ 39.23 kPa

At 0% level:

LRV ≈ -39.23 kPa

The 3.0 m process level at 800 kg/m³ produces:

pprocess,100% = 800 × 9.80665 × 3.0

pprocess,100% ≈ 23.54 kPa

At 100%, the differential pressure is therefore:

URV = 23.54 - 39.23

URV ≈ -15.69 kPa

Condition Process column Wet-leg pressure Differential pressure Output
0% level 0 kPa 39.23 kPa −39.23 kPa 4 mA
50% level approx. 11.77 kPa 39.23 kPa approx. −27.46 kPa approx. 12 mA
100% level 23.54 kPa 39.23 kPa −15.69 kPa 20 mA

The entire calibrated differential pressure range in this example lies in the negative range. Nevertheless, as the level increases, the differential pressure continuously rises from −39.23 to −15.69 kPa.

The control system can easily scale this range to 0 to 100% or 4 to 20 mA.

Influence of process and reference density

In hydrostatic level measurement, density is not a secondary parameter. The transmitter measures pressure, not geometric height directly.

If the process density changes while the actual level remains constant, the hydrostatic pressure on the H side changes. The indication can therefore change even though the geometric liquid height has not.

With a wet leg, there is also a second density: the density of the reference liquid.

This means that two independent density changes can affect the measuring point:

  • If ρP changes, the slope of the level characteristic changes.
  • If ρW changes, the reference pressure or zero point of the measurement shifts primarily.

This is very useful for diagnostics. A change in process density typically has a level-dependent effect. A changed wet-leg density, by contrast, produces a largely constant additional offset.

With media whose density depends strongly on temperature or concentration, it should therefore be established during the design phase whether the required accuracy can be achieved using a fixed density assumption or whether additional density or temperature compensation is required.

Why the temperature of the reference line matters

The wet-leg liquid also has a temperature-dependent density.

If, for example, a water-filled reference line warms up, the liquid density decreases. The hydrostatic pressure column on the low-pressure side therefore becomes smaller.

With the usual connection arrangement, H at the bottom and L at the wet leg, this means:

The L-side pressure decreases, the measured differential pressure H - L increases and the indicated level may appear too high.

If the reference column cools and its density increases, the effect can reverse.

A long reference line passing through several temperature zones is particularly critical. One section may be located in a heated area while another is outdoors. In that case, the effective mean density can no longer be derived simply from one single room temperature.

For demanding applications, ambient temperature, insulation and line routing therefore form part of the measuring-point design.

One-sided heat tracing or a strong local heat source can influence the reference column just as direct sunlight can.

Why the wet-leg liquid level must remain constant

The calculation is reproducible only if the vertical liquid column on the L side is known.

If the liquid level in the wet leg drops, its hydrostatic counterpressure becomes smaller. The measured differential pressure rises and the transmitter indicates a higher level even though the process itself has not changed.

The same applies to a larger gas bubble in the reference line. Where the gas bubble is located, part of the intended liquid column is missing. The effective hydrostatic reference pressure decreases.

Conversely, an additional denser fill liquid or a higher reference level can increase the L-side pressure and shift the level indication downwards.

For reliable measurement, it is therefore not enough for the reference line to be filled correctly only during initial start-up. Its condition must remain reproducible throughout operation.

Function of the condensate pot

In vessels containing condensing vapour, a condensate pot or condensate chamber is often installed at the upper reference connection.

The vessel establishes a defined upper reference point for the liquid column. Condensing vapour can form liquid there and the connected reference line is filled with condensate in a controlled manner.

The vertical height between this reference level and the transmitter, together with the liquid density, determines the hydrostatic pressure on the L side.

For the concept to work, the installation height and filling condition of the condensate pot must match the calculation.

A shifted vessel, a later change in transmitter height or a line that is subsequently rerouted partly higher or lower can change the calculated reference pressure.

During maintenance, therefore, not only the transmitter itself should be documented. The condensate pot, impulse line and installation heights also belong to the measuring point.

Include transmitter height and impulse lines in the calculation

For simplicity, the equation used so far assumed that the transmitter was installed at the same height as the lower process connection.

In real installations, this is often not the case.

If the differential pressure transmitter is, for example, installed one metre below the lower process connection and the H-side line is filled with process liquid, this liquid column also generates an additional constant pressure.

A more general expression is therefore:

Δp = ρP × g × h + ρH × g × HH - ρW × g × HW

Here, ρH × g × HH describes the constant hydrostatic contribution of the high-pressure impulse line.

The actual vertical heights must therefore be used for parameterisation. The total pipe or hose length is not decisive for the static liquid head.

A ten-metre-long horizontally routed line does not generate the same hydrostatic shift as a ten-metre-high liquid column.

After any later change to the installation height of the transmitter, LRV and URV should therefore be checked again.

Correctly parameterising the 4–20 mA range

A common misunderstanding is to assume that 4 mA must always correspond to 0 mbar differential pressure.

With wet-leg level measurement, this is often specifically not the case.

In the example calculated above:

−39.23 kPa = 4 mA = 0%

and:

−15.69 kPa = 20 mA = 100%

The transmitter is therefore not set to “0 to 23.54 kPa”, but to the actual differential pressure range present in the process.

The control system then receives a normal linear 4–20 mA signal and can calculate percentage, level height or – using a tank characteristic – volume.

With nonlinear tank geometries such as horizontal cylinders or spherical vessels, level height should not simply be interpreted linearly as volume. Appropriate linearisation or a tank table is required.

Distinguishing zero adjustment from zero shift

The term “zero” often causes misunderstandings in wet-leg applications.

Using a valve manifold, a differential pressure transmitter can be subjected to the same pressure on the H and L sides. The physical differential pressure at the measuring cell is then 0 mbar. This condition can be used for checking or for instrument zero adjustment.

However, this does not mean that 0 mbar in the real process corresponds to 0% level.

After returning to normal process conditions, the wet-leg liquid column again acts on the L side. With an empty tank, the calculated negative differential pressure therefore appears again.

Two different conditions must therefore be clearly distinguished:

  • Sensor zero: Both inputs are subjected to the same pressure, Δp = 0.
  • Process zero or 0% level: The actual differential pressure at minimum level including all hydrostatic installation effects.

If these two conditions are confused, a correctly designed wet-leg measuring point can be completely mis-scaled by an incorrect zero adjustment.

Correctly scaling alarms and the control system

For process control, the raw differential pressure value is not normally used directly. Instead, the scaled level derived from it is used.

Limit values such as:

  • Low,
  • Low-Low,
  • High,
  • High-High

should therefore be based on a correctly parameterised and documented level scale.

For maintenance purposes, however, it is useful to keep the raw differential pressure available as well. An apparently drifting level is much easier to diagnose if it is known whether the actual Δp raw value has changed.

For critical applications, the temperature of the reference line or condensate area can also be useful as an additional diagnostic variable. This can help distinguish a temperature-induced zero shift from a real level change.

Continuous level measurement and independent point-level detection perform different functions. Whether a separate overfill or dry-run protection system is required for a particular installation depends on the respective plant and safety concept.

Recognising typical wet-leg faults

Observation Possible cause Sensible check
Indicated level rises slowly without actual filling Wet leg is losing liquid or contains gas Check fill condition, condensate pot and reference line
Level changes with outside temperature Density change of the wet-leg liquid Check temperature trend and line routing
Constant offset since commissioning LRV/URV or installation height calculated incorrectly Recalculate geometry, densities and parameterisation
Measured value jumps after maintenance work Reference line not completely filled or not properly vented Restore the wet leg according to the operating procedure
Measured value is correct only at one particular product temperature Process or reference density not correctly considered Check density-temperature behaviour
Transmitter correctly shows 0 mbar at the valve manifold, but process indication is still wrong Instrument zero is correct, but application scaling or wet-leg calculation is incorrect Check LRV, URV and process geometry

These fault patterns show why differential pressure level measurement must not be assessed solely by looking at the electronic instrument.

The transmitter, valve manifold, both impulse lines, condensate pot, fill media, densities and installation heights together form the complete measuring chain.

When diaphragm seals or electronic differential pressure measurement are useful

A conventional wet leg is proven, easy to understand and highly reliable in many plants. However, it requires a defined and permanently stable liquid column.

For difficult media or high temperatures, diaphragm seal systems can be an alternative. In this arrangement, diaphragms separate the process from the measuring system and pressure is transmitted through completely filled capillary systems.

This eliminates the classic open wet-leg impulse line. At the same time, however, other influences arise, particularly from fill fluid, capillary length, installation height and ambient temperature.

With very large vertical distances, an electronic differential pressure solution may also be attractive. In this case, two pressure transmitters measure the two process pressures separately and the difference is calculated electronically.

Such a solution does not use a shared reference liquid column several metres high. However, the accuracy of the two separate pressure measurements and the static process pressure must be suitable for the required small differential pressure.

There is therefore no single fundamentally best solution. Wet leg, diaphragm seals and electronic differential pressure measurement must be compared based on medium, temperature, pressure, height, maintenance, measuring span and required accuracy.

Systematic design and commissioning procedure

For reliable wet-leg level measurement, the complete measuring point should be described geometrically and hydraulically before parameterisation.

  1. Define the measuring range: Clearly define the lower and upper level.
  2. Determine vessel pressure: Record the minimum and maximum static gas-space pressure.
  3. Determine process density: Document density and, where applicable, its dependence on temperature or concentration.
  4. Select the reference type: Deliberately choose dry leg or wet leg.
  5. Define the wet-leg liquid: Specify condensate or another suitable fill medium.
  6. Determine wet-leg density: Consider the relevant temperature condition.
  7. Record vertical heights: Precisely document process connections, condensate pot and transmitter position.
  8. Consider the high-pressure impulse line: Include any additional liquid column between the lower connection and transmitter.
  9. Calculate LRV: Determine differential pressure at minimum level.
  10. Calculate URV: Determine differential pressure at maximum level.
  11. Check transmitter range: Negative and positive range limits must remain within the instrument specification.
  12. Fill the reference line completely: Remove air and vapour bubbles according to the specified commissioning procedure.
  13. Check valve manifold and connections: Clearly assign H and L sides.
  14. Set the 4–20 mA scaling: Parameterise LRV = 4 mA and URV = 20 mA.
  15. Check process plausibility: Compare indicated level with a known plant condition.
  16. Document the initial condition: Record densities, heights, temperatures and configuration for later maintenance.

Documenting the heights and densities is particularly important. If the transmitter is replaced years later, the information “measuring range 0 to 100%” is not sufficient to reconstruct the wet-leg configuration correctly.

Common planning and commissioning errors

Equating 0% level with 0 mbar

In a wet-leg arrangement, a significantly negative differential pressure is often present with an empty vessel. This value is part of the correct measuring-point calculation.

Equating wet-leg density with process density

The reference liquid and process medium may have different densities and different temperatures. Both must be considered separately.

Using total line length instead of vertical height

Hydrostatic pressure depends on the vertical liquid column, not on the total geometric pipe length.

Accepting a partially filled reference line

A changing liquid level creates a changing reference pressure and therefore a direct level error.

Programming 0 mbar as the 4 mA point after zero adjustment

Instrument zero and process zero are different conditions in wet-leg measurement.

Changing the transmitter height afterwards

This changes the vertical liquid columns in the impulse lines. LRV and URV must then be checked again.

Ignoring temperature changes in the wet-leg liquid

The density of the reference liquid changes with temperature. With long wet legs and small spans, this can produce a relevant zero shift.

Including gas-space pressure in the level span

With correct differential pressure connection, the gas-space pressure acts on both sides and largely cancels out. The hydrostatic differences are decisive.

Calibrating only the differential pressure transmitter

A correctly operating instrument does not guarantee correct level measurement. Impulse lines, reference column, densities and geometry also form part of the measuring chain.

Suitable measurement technology at ICS Schneider

ICS Schneider Messtechnik offers differential pressure and level measurement technology for open and closed process vessels as well as demanding industrial applications.

The Siemens SITRANS P320 is suitable for gauge pressure, absolute pressure, differential pressure, flow and level measurement. It can therefore be used for classic level measurement in closed vessels with H/L process connections.

In wet-leg applications, the actual differential pressure range is determined from process density, reference column and installation geometry and is then parameterised in the transmitter as the lower and upper range values.

The WIKA DPT-20 is also designed for industrial differential pressure and level measurement in open and closed tanks. Different measuring ranges and process versions allow adaptation to small differential pressures combined with high static process pressure.

For difficult process conditions, diaphragm seal solutions and suitable accessories for differential pressure measuring points are also available.

Level measurement technology at ICS Schneider

Process and differential pressure transmitters at ICS Schneider

Further reading: Measuring level in a pressurised tank – correctly compensating gas-space pressure

Further reading: Differential pressure transmitter provides implausible values

Conclusion

A wet reference leg makes differential pressure level measurement in closed vessels with a condensing gas or vapour space reproducible. The low-pressure side is deliberately subjected to a defined liquid column.

This reference column is not an unwanted interfering pressure, but part of the measuring-point calculation. Its height and density shift the entire differential pressure range and often result in both 0% and 100% level corresponding to negative differential pressures.

The process density essentially determines the usable level span. The density and height of the wet leg, by contrast, determine a significant part of the constant offset. In real installations, the transmitter mounting height and liquid columns in the impulse lines must also be considered.

Temperature deserves particular attention. If the density of the reference liquid changes, the hydrostatic pressure on the L side changes and therefore so does the level indication. A wet leg must therefore not only remain completely filled, but also be operated under thermally and geometrically reproducible conditions.

Anyone who calculates LRV and URV from the actual plant geometry, clearly distinguishes instrument zero from process zero, and considers wet-leg fill level, density and temperature during maintenance will obtain a robust and traceable differential pressure level measurement.

FAQ on level measurement with a wet leg

What is a wet leg in differential pressure level measurement?

A wet leg is a completely liquid-filled reference line on the low-pressure side of a differential pressure transmitter. It creates a defined hydrostatic reference column.

When is a wet reference leg used?

It is used particularly when condensation is expected in the gas or vapour space and a permanently dry reference line therefore cannot be guaranteed.

Which side of the differential pressure transmitter is connected to the wet leg?

In classic level measurement in a closed vessel, the lower process connection is connected to H or +, while the wet reference line from the gas space is connected to L or −.

Why is the differential pressure often negative when the tank is empty?

The filled reference line generates hydrostatic pressure on the low-pressure side. If this is greater than the pressure on the high-pressure side at minimum level, the resulting differential pressure is negative.

Is a negative differential pressure at 0% a fault?

No. In a correctly calculated wet-leg arrangement, this is completely normal. The key is to parameterise LRV and URV accordingly.

What is zero elevation?

This term is often used for a range shift in which 0% of the process value corresponds to a negative differential pressure. Wet-leg applications are a typical example.

How is wet-leg pressure calculated?

In simplified form using p = ρ × g × h. The density of the reference liquid and the vertical height of the liquid column are required.

Which density must be used for the wet leg?

The density of the actual reference liquid under the relevant temperature conditions. It does not have to be identical to the density of the process medium.

Does the temperature of the wet leg affect the level?

Yes. The density of the reference liquid changes with temperature. This changes the hydrostatic pressure on the L side and can shift the zero point of the level measurement.

What happens if the wet-leg line becomes partially empty?

The hydrostatic counterpressure on the low-pressure side decreases. With the usual connection arrangement, the measured differential pressure therefore increases and the indicated level may be too high.

Can gas bubbles in the wet leg cause a measurement error?

Yes. They reduce the effective liquid column and can therefore cause an offset or unstable measured values.

Why is a condensate pot used?

With condensing vapours, it establishes a defined upper reference point and supports a reproducible, completely filled reference column.

Does a wet leg change the actual measuring span?

With a constant reference column, the wet leg mainly shifts the measuring range. The span generated by the change in process level remains fundamentally determined by process density and the level range.

Can 4 mA correspond to a negative differential pressure?

Yes. A differential pressure transmitter can, for example, be parameterised so that −400 mbar corresponds to 4 mA or 0%, and −150 mbar corresponds to 20 mA or 100%.

Is zero adjustment with the equalising valve open the same as 0% level?

No. During equalisation, H and L are subjected to the same pressure and the measuring cell sees 0 mbar. The actual process condition at 0% can have a significantly negative differential pressure because of the wet leg.

Does the transmitter height have to be taken into account?

Yes. Vertical liquid columns between the process connections and transmitter create additional constant hydrostatic pressure contributions and must be included in the LRV/URV calculation.

Is the total length of the impulse line relevant to hydrostatic pressure?

No. For the static hydrostatic contribution, the vertical height difference is the main factor, not the total line length.

When is a diaphragm seal system an alternative to a wet leg?

Diaphragm seals can be useful when impulse lines are problematic because of the medium, temperature, viscosity, crystallisation or maintenance requirements. However, they introduce their own temperature and capillary effects.

Which transmitter is suitable for wet-leg level measurement?

Suitable industrial differential pressure transmitters such as the Siemens SITRANS P320 or WIKA DPT-20 can be used for these applications, provided the measuring range, static pressure, process connections and instrument version match the specific installation.

Diese Website benutzt Cookies. Wenn du die Website weiter nutzt, gehen wir von deinem Einverständnis aus.