Tank Level Measurement with DP Transmitters

Hydrostatic Pressure and Specific Gravity

Differential pressure transmitter level measurement is an indirect method. The transmitter does not “see” the liquid surface like a radar, ultrasonic, or float device. It measures the pressure created by the liquid column above a fixed point, then converts that hydrostatic pressure into level.

The basic relationship is:

\[ P = \rho g h \]

Where:

\(P\) = hydrostatic pressure
\(\rho\) = liquid density
\(g\) = acceleration due to gravity
\(h\) = height of the liquid column above the measuring point

In field calibration, technicians often use equivalent head units, such as inches of water column or millimeters of water column. When pressure is expressed as an equivalent water column, the common relationship is:

\[ DP = \text{liquid height} \times SG \]

Therefore:

\[ \text{liquid height} = \frac{DP}{SG} \]

This assumes DP is expressed in the same equivalent head units as the level height and that specific gravity, or SG, is dimensionless. For example, if the transmitter reads 80 inches of water column and the process liquid has an SG of 0.8, the corresponding liquid height is:

\[ \frac{80}{0.8} = 100 \text{ inches of liquid} \]

Specific gravity is essential because equal heights of different liquids do not produce equal pressures. Water with SG 1.0 produces more pressure than oil with SG 0.8 at the same height. A heavier liquid produces more pressure for the same height. For example, 100 inches of a process liquid with SG 1.2 corresponds to:

\[ 100 \times 1.2 = 120 \text{ inches of water column} \]

This is why vessel height alone is not enough for calibration. If a transmitter is ranged for 100 inches of water but the tank contains a liquid with SG 1.2, the actual full-scale pressure is 120 inches of water column. Ignoring that conversion creates level error.

For the pressure-to-level conversion to be valid, several assumptions must be reasonably true:

The liquid density is constant or changes only within an acceptable range.
The SG used for calibration matches the liquid at the relevant operating temperature.
The reference pressure is stable and correctly applied to the low-pressure side.
The measurement tap is fixed relative to the required level range.
Tank conditions do not introduce major pressure gradients unrelated to liquid height.

For a constant-density liquid in normal tank conditions, bottom pressure rises approximately linearly with liquid level. Doubling the height doubles the hydrostatic head. This linear relationship is why DP transmitters are widely used for tank level applications: once density and geometry are known, output can be ranged directly to level.

The limitation is that the transmitter measures pressure, not volume or surface position independent of density. If the liquid changes composition, temperature, or density, the same DP may represent a different actual level. Where SG varies significantly, compensation or another level technology may be required.

Open-Tank Level Measurement Setup

Open tanks are one of the simplest applications for differential pressure transmitter level measurement because the vapor space above the liquid is exposed to atmosphere. There is no closed gas pressure that must be removed from the measurement.

In a typical open-tank arrangement, the high-pressure side of the transmitter is connected near the lowest measuring point on the tank. This connection senses hydrostatic pressure from the liquid above it. The tap is usually at or below the 0% level reference, depending on the mechanical arrangement and calibrated range.

The low-pressure side is vented or left open to atmosphere. Since the liquid surface is also exposed to atmosphere, the atmospheric pressure acting on the surface and the atmospheric reference at the transmitter cancel. The transmitter effectively measures bottom gauge pressure.

In simplified terms:

\[ DP = P_{bottom} - P_{atmosphere} \]

For an open tank:

\[ P_{bottom} = P_{atmosphere} + P_{liquid head} \]

So:

\[ DP = P_{liquid head} \]

The resulting differential pressure is proportional to liquid height above the high-pressure tap, provided the liquid density or SG is known and reasonably constant.

A practical open-tank measurement chain looks like this:

Liquid level increases.
Hydrostatic pressure at the bottom tap increases.
The transmitter senses higher pressure on the high-pressure side.
The DP signal increases.
The output increases toward the upper range value.

When level falls, bottom pressure decreases, DP decreases, and output moves toward the lower range value.

The main calibration task is converting the level range into pressure. If the tank level span is 0 to 100 inches and the liquid SG is 1.0, the pressure span is 0 to 100 inches of water column. With SG 0.8, the span is 0 to 80 inches of water column. With SG 1.2, it is 0 to 120 inches of water column.

This configuration is straightforward, but it still requires good installation practice. The pressure tap should represent the intended measuring point, impulse lines should avoid plugging or trapped gas where that affects measurement, and the transmitter should be ranged for the actual process liquid. An open tank simplifies the reference pressure; it does not remove density correction.

Closed or Pressurized Tank Arrangement

Closed and pressurized tanks require a different arrangement because bottom pressure includes both liquid head and the pressure of the gas or vapor space above the liquid. If only bottom pressure is measured, any vapor-pressure change appears as a false level change.

For a closed tank:

\[ P_{bottom} = P_{vapor} + P_{liquid head} \]

If a transmitter were connected only to the bottom, a rise in vapor pressure would increase the reading even if level did not change. A drop in vapor pressure would decrease the reading even if level remained constant.

To compensate, the low-pressure side of the differential pressure transmitter is connected to the top vapor space. The high-pressure side remains connected near the bottom measuring point. The transmitter subtracts vapor-space pressure from bottom pressure:

\[ DP = P_{bottom} - P_{vapor} \]

Substituting the bottom pressure expression:

\[ DP = (P_{vapor} + P_{liquid head}) - P_{vapor} \]

So:

\[ DP = P_{liquid head} \]

Under intended conditions, the remaining differential pressure corresponds mainly to the liquid column. This is why DP transmitters are useful on pressurized vessels: they can reject common pressure that appears on both top and bottom connections.

This compensation works only when the top pressure is accurately transmitted to the low-pressure side. The impulse line, remote seal, or connection method must represent actual vapor-space pressure without unwanted liquid accumulation, blockage, leakage, or temperature effects that create additional pressure.

A dry-leg installation is used when the low-pressure impulse line contains only gas or vapor and remains dry. It is appropriate only where the vapor does not condense under expected operating and ambient conditions. In a dry leg, the low side transmits vapor-space pressure without a standing liquid column.

A dry-leg closed-tank setup is conceptually simple:

High-pressure side: connected to the lower tank tap.
Low-pressure side: connected to the vapor space.
The transmitter subtracts vapor pressure from bottom pressure.
The result is hydrostatic pressure from liquid level.

The limitation is that many vapors do not remain dry. Temperature changes, heat loss in impulse tubing, or process cycling can cause condensation. Once liquid collects in a line assumed to be dry, the transmitter sees added hydrostatic head from that liquid. Condensing services often require a wet-leg arrangement instead.

Wet-Leg Impulse Lines for Condensing Vapor Service

Condensing vapor service can make a dry low-pressure impulse line unreliable. In applications such as steam boilers, heated vessels, or tanks containing condensable vapors, vapor in the low-pressure line can cool and form liquid. If condensate amount changes over time, the pressure applied to the transmitter low side also changes.

This is a serious measurement problem because the transmitter interprets any pressure change between high and low sides as a level change. If condensate accumulates unevenly in the low-pressure impulse line, indicated level may drift, jump, or become unstable while actual tank level is steady.

For example, consider a closed vessel with a dry-leg connection to the vapor space. If the low-pressure line starts dry, the transmitter subtracts only vapor pressure. Later, vapor condenses and creates a partial liquid column in that line. The hydrostatic pressure from that liquid is now applied to the low-pressure side. Since the transmitter calculates:

\[ DP = HP - LP \]

an increase in LP-side pressure reduces measured DP. The transmitter may indicate a lower level than actual. If condensate level in the impulse line changes irregularly, the error also changes irregularly.

A wet leg makes the reference condition intentional instead of accidental. It is a low-pressure impulse line deliberately filled with a stable liquid. Rather than hoping the line remains dry, the installation maintains a predictable reference column on the low-pressure side.

In condensing vapor applications, the wet leg provides a consistent hydrostatic pressure. Because that pressure is known or can be included during calibration, the transmitter range can be set correctly. The goal is not to eliminate low-side hydrostatic pressure, but to make it stable and repeatable.

The wet-leg fill or isolation fluid must suit the service. It should remain stable under operating conditions, be compatible with the process and wetted materials, and be appropriate for expected temperature and pressure. It should not react with the process, degrade rapidly, vaporize unexpectedly, freeze under ambient conditions, or create maintenance problems. Specific selection depends on the process and plant practice.

Wet legs are common where condensation is expected or unavoidable. Steam boiler drum level measurement is a typical example because steam can condense in impulse piping. Other applications include vessels with hot condensable vapors, reactors with vapor recovery systems, or services where ambient cooling of the impulse line can form liquid.

A wet leg improves repeatability, but it changes calibration. The low-pressure side now has a standing hydrostatic head even when vessel level is at 0%. That fixed pressure must be included in LRV and URV calculations.

Correcting the Wet-Leg Zero Offset

A wet leg applies constant hydrostatic pressure to the transmitter low-pressure side. This pressure is not caused by process level inside the tank; it is caused by the known liquid column in the low-pressure impulse line. Because the transmitter measures the difference between high and low pressure, this reference column creates an offset.

Depending on tank tap elevations, transmitter location, and wet-leg fill-fluid density, the low-pressure side can apply more pressure than the high-pressure side when the vessel is at 0% level. In that case, the transmitter sees negative differential pressure at the empty or zero-level condition.

For example, at 0% level:

\[ DP = HP - LP \]

If the high-pressure side has little or no process liquid head, but the low-pressure side has a full wet-leg column, then:

\[ LP > HP \]

So:

\[ DP < 0 \]

This does not mean the transmitter is faulty. It means the installation creates a real differential pressure below zero on the HP-minus-LP scale. Calibration must account for it.

The practical correction is to treat wet-leg back pressure as the 0% level reference. The transmitter LRV is not automatically zero pressure; it is the actual differential pressure that exists at the defined 0% level. If that value is negative, the transmitter should be configured so the negative DP corresponds to 4 mA, or to 0% level in the digital output.

Terminology can vary. Some sources distinguish zero suppression, zero elevation, and zero migration based on offset sign and installation geometry. Rather than relying only on a label, calculate the actual DP at 0% and 100% level, including wet-leg pressure, and assign the transmitter range accordingly.

Without this correction, indicated level may remain below zero, fail to reach the correct full-scale value, or otherwise not match actual tank level. A wet leg improves stability in condensing vapor service only if the constant low-side head is included in calibration.

A simplified way to think about wet-leg calibration is:

The wet leg creates constant pressure on the LP side.
The process liquid creates variable pressure on the HP side.
The transmitter output must be based on actual DP at the lower and upper level limits.
The 0% point is calculated from geometry and fluid densities.

The correction also depends on maintaining the wet leg as intended. If the fill level changes, leaks, flashes, or becomes contaminated, the assumed low-side pressure changes. That changes the transmitter zero reference and produces level error. Wet-leg systems therefore require attention to installation details, isolation, filling, and maintenance access.

Setting LRV and URV for Calibration

After mechanical installation, the transmitter range is configured by assigning a lower range value and an upper range value. These values define how measured differential pressure maps to transmitter output.

LRV means lower range value. It is the differential pressure corresponding to the 0% level condition. In a conventional 4-20 mA loop, the LRV is mapped to 4 mA.

URV means upper range value. It is the differential pressure corresponding to the 100% level condition. In a conventional 4-20 mA loop, the URV is mapped to 20 mA.

For level measurement, LRV and URV should be assigned in pressure units, not merely tank height. Vessel height must first be converted into pressure or equivalent water column using liquid SG.

For a simple open tank with the transmitter high-pressure side connected at the 0% reference elevation and the low-pressure side vented to atmosphere:

\[ LRV = 0 \]\[ URV = \text{level span} \times SG \]

If the level span is 100 inches and the liquid SG is 1.2:

\[ URV = 100 \times 1.2 = 120 \text{ inches of water column} \]

So the transmitter would be ranged from 0 to 120 inches of water column for a 0 to 100 inch level span. The 4 mA signal represents 0 inches of level, and the 20 mA signal represents 100 inches.

For a liquid with SG 0.8 over the same 100 inch span:

\[ URV = 100 \times 0.8 = 80 \text{ inches of water column} \]

The physical height range is the same, but the pressure range is different. This is why SG conversion must be performed before calibration.

For closed tanks, the same hydrostatic conversion applies, but the vapor-space reference must be included. In a dry-leg arrangement where the low-pressure side accurately tracks vapor pressure and no liquid column exists in the low-side line, the transmitter range can often be based mainly on liquid head between the lower and upper level limits. Top pressure is common to both sides and is subtracted by the transmitter.

For wet-leg arrangements, LRV and URV must include the constant low-pressure-side hydrostatic head. A general approach is:

Determine the HP-side pressure at 0% level.
Determine the LP-side pressure from the wet leg at 0% level.
Calculate \(LRV = HP_{0\%} - LP_{0\%}\).
Determine the HP-side pressure at 100% level.
Use the same stable wet-leg reference pressure, unless the installation geometry creates another known condition.
Calculate \(URV = HP_{100\%} - LP_{100\%}\).

In many wet-leg installations, LRV may be negative because low-side wet-leg pressure is greater than high-side pressure at 0% level. URV may still be negative, zero, or positive depending on process span, wet-leg height, and fill-fluid density. The correct range is the actual DP range created by the installed system.

Common calibration errors include:

Entering vessel height as pressure without multiplying by SG.
Assuming water calibration values apply to a lighter or heavier process liquid.
Ignoring wet-leg pressure on the low-pressure side.
Treating a closed tank like an open tank and failing to compensate for vapor pressure.
Assuming a dry leg will remain dry in condensing service.
Using SG at ambient conditions when operating temperature significantly changes density.

The central rule is simple: configure the transmitter for the differential pressure it will actually see at 0% and 100% level. For reliable differential pressure transmitter level measurement, those values must reflect liquid SG, reference pressure, installation geometry, and any impulse-line liquid columns.