Problem Summary
The SITRANS P320 hydrostatic pressure transmitter configured for PARAB (parabolic bottom) volume measurement is failing to track level in the lower quartile of a 1.22 m parabolic-bottom vessel. The transmitter was set with the following parameters:
| Parameter (Index) | Name | Configured Value |
|---|---|---|
| [2] | Lower Range Value (LRV) | 0 mbar |
| [3] | Upper Range Value (URV) | 120 mbar |
| [5] | Application | PARAB |
| [18] | Lower Scaling Point | 0 L |
| [19] | Upper Scaling Point | 1112 L |
| [21] | Vessel Dimension A | 7.01 % (entered as 78 L) |
| [22] | Vessel Dimension L | (per configuration, not yet set) |
Field observation: the 4–20 mA output and digital HART primary variable remain at zero (4.000 mA) until the physical level exceeds roughly 25 % of the tank height. Above 25 % the reading begins to climb, but the indicated volume is no longer linear with the actual hydrostatic head. The expected behavior is a linear 0–1112 L output from 0 % to 100 % fill with the transmitter installed at the lowest practical elevation on the vessel.
Hydrostatic Pressure Verification
Before touching any transmitter parameter, verify that the LRV/URV pair matches the actual hydrostatic column at the diaphragm. For a water-filled vessel at rest with a vented reference (gauge measurement):
P = ρ · g · h
Where:
- ρ = density of process fluid (1000 kg/m³ for water at 4–20 °C)
- g = 9.81 m/s² (local gravitational acceleration; adjust for altitude if > 1000 m)
- h = vertical liquid height above the diaphragm in meters
For the reference installation (h = 1.22 m, water):
P = 1000 · 9.81 · 1.22 = 11 968 Pa ≈ 119.68 mbar
This matches the configured URV of 120 mbar within instrument tolerance, so the LRV/URV pair is dimensionally correct. The user’s 120 mbar figure is right.
Mounting Location on a Parabolic-Bottom Vessel
Per the SITRANS P320/P420 HART Operating Instructions (A5E44852162, Siemens Industry Online Support), the mounting location must satisfy several conditions: accessibility, proximity to the measuring point, ambient temperature within the specified limits, and protection from vibration and contamination. For a hydrostatic level application the additional, application-specific requirement is that the transmitter be installed at the lowest practical elevation on the vessel.
A pressure transmitter measures only the hydrostatic head of liquid above its sensing diaphragm. Any liquid column below the elevation of the process connection does not contribute to the measured pressure and will appear as 0 % level (4.000 mA). On a vertical tank with a parabolic (or dished) bottom, the geometric low point is at the apex of the parabola, but a process nozzle is almost always located on a flat section of the head or on a sidewall above the apex. This unavoidable offset creates a fixed dead band equal to the vertical distance between the diaphragm and the geometric low point.
Recommended Mounting Position for PARAB
- Mount the transmitter on the bottom head of the vessel, as close to the apex of the parabolic section as the nozzle geometry allows. The diaphragm must be the lowest point of the wetted measurement column.
- Avoid sidewall mounting above the parabolic section. A sidewall tap loses the entire parabolic head volume, and the indicated level will not begin to rise until the liquid reaches the elevation of the tap.
- Do not mount on the top head or in the vapor space. The transmitter will read negative gauge pressure (or vacuum) when the vessel is empty, which is outside the calibrated range.
- Use a siphon or impulse line only if the process temperature exceeds the sensor limit (typically 85 °C for the standard SITRANS P320 fill fluid). The siphon must be liquid-filled at all times.
- Verify that the process connection does not protrude into the tank such that liquid is trapped under the diaphragm. Air or gas pockets at the diaphragm will cause a static offset and slow response.
For the reference 1.22 m parabolic vessel, the diaphragm should be at the bottom apex (or as close to it as the nozzle permits). If the nozzle is, for example, 21 mm above the apex, the first 21 mm of fill will not be measured. The installed elevation must be entered into the geometry calculation so that Dimension L accounts for the total cylindrical section height above the diaphragm, not above the tank low point.
PARAB vs PARAE Application Modes
The SITRANS P320 exposes several tank-shape profiles in the Application parameter [5]. The two parabolic modes are:
| Mode | Geometry | Use Case |
|---|---|---|
| PARAB | Vertical cylinder with parabolic (domed) bottom | Vertical tank with a single dished head on the bottom; flat or open top |
| PARAE | Horizontal cylinder with parabolic (domed) ends on both sides | Horizontal tank, or any geometry with two dished heads contributing volume |
Both modes use the same internal arithmetic structure:
Volume ∝ Dimension L + (2 · Dimension A)
The factor of 2 reflects the fact that the parabolic profile has two identical contributing faces in the formula derivation. For a horizontal tank with two dished ends, this is physically correct: there are two heads. For a vertical tank with a single dished bottom, the factor of 2 is either (a) compensated inside the firmware with a built-in halving, (b) a manual entry convention where Dimension A is already the half-height of the parabola, or (c) a documentation/copy error that requires the user to divide their measured parabola height by 2 before entry. The Operating Instructions do not unambiguously state which interpretation is in force.
Vessel Dimension A: Units and Interpretation
Vessel Dimension A [21] is the parameter that defines the depth (or height) of the parabolic head. The published convention is that A is entered as a percentage of the total vessel height, expressed in the same engineering units the operator has selected for the scaled output (in this case, liters).
Re-deriving the Correct A Value
For the reference vessel:
- Total vessel height: 1220 mm
- Parabolic section height (apex to start of straight cylindrical wall): 21 mm
- Straight cylindrical section height: 1220 − 21 = 1199 mm
If A is entered as a percentage of total height:
A = (21 / 1220) · 100 = 1.72 %
The user has entered 7.01 %, which is 4× the correct value. This single error shifts the volume model so that the lower quartile of physical fill (0–25 %) is mapped into a “dead zone” of the calculation, producing exactly the observed symptom: the transmitter output does not begin to climb until the tank is roughly one-quarter full.
| Entry in [21] | Interpretation | Observed Effect |
|---|---|---|
| 78 (raw integer, no %) | Probably rejected as out-of-range or interpreted as 78 % | Severe non-linearity, possible clamp at LRV |
| 7.01 (no %) | 7.01 % if firmware strips % sign; or 7.01 raw units if not | Volume tracks above ~25 % fill, dead band below |
| 1.7 (%) | Correct geometric interpretation of the parabola as a % of total height | Volume tracks from 0 % fill through 100 % fill |
| 21 (mm or %) | If firmware treats numeric 21 as 21 %, dead band extends to ~21 % | Output flat below 21 % fill, climbs thereafter |
The pattern “output stays at zero until the tank is ~X % full, then begins to climb linearly” is the diagnostic signature of an A-dimension entered in the wrong units. The break point of the output curve corresponds to the A value as a percentage of total volume.
Vessel Dimension L and the Straight Section
Vessel Dimension L [22] is the length of the straight cylindrical (or straight-walled) section of the vessel. For a vertical tank with a parabolic bottom, L is the height from the top of the parabolic section to the overflow point or to the physical top of the straight wall, in the same engineering units as A.
For the reference installation, if the firmware treats L and A as percentages:
- L = (1199 / 1220) · 100 = 98.28 %
- A = (21 / 1220) · 100 = 1.72 %
- Sum: L + A = 100 %, with the 2·A contribution negligible in this geometry.
If the firmware treats L and A as raw engineering units (mm in this case, but the firmware will internally scale against LRV/URV), then:
- L = 1199 mm
- A = 21 mm (full parabola height, or 10.5 mm if the factor-of-2 compensation is required by user entry)
The commissioning procedure must determine which unit convention is in force on the actual installed firmware revision (read the device tag or use HART identify). The Operating Instructions manual should be cross-checked against the unit displayed in SIMATIC PDM or the local LCD for parameter [21].
Step-by-Step Resolution Procedure
- Confirm the physical installation. Verify that the diaphragm is at the lowest practical point on the vessel. Measure the offset between the diaphragm and the geometric low point of the parabolic section in millimetres; this is the irreducible dead band and must be documented.
- Confirm the HART address and revision. From a HART handheld or SIMATIC PDM, read the device identification. Note firmware version, hardware revision, and device tag. The current revision of the operating instructions is A5E44852162; consult the manual revision-history table to map parameter behaviour to the installed firmware.
- Set Application to PARAB [5]. Confirm this is the active value. If the vessel is actually horizontal with two dished ends, switch to PARAE. The geometry, dimension names, and formula structure are different; do not assume cross-compatibility.
- Set LRV [2] and URV [3] to the calibrated hydrostatic range. For water at 1.22 m column: 0 mbar to 120 mbar (rounded up from 119.7 mbar to give 0.3 % headroom).
- Set Lower Scaling [18] and Upper Scaling [19] to the engineering range. 0 L to 1112 L in this case, matching the desired 4–20 mA endpoints.
- Determine the A-dimension unit convention. Read the on-device LCD or SIMATIC PDM display for parameter [21] while the value is highlighted. The unit symbol (%, mm, m, ft, in, L, gal, etc.) is shown next to the numeric value. This single observation resolves the documentation ambiguity for the installed firmware.
- Enter Dimension A [21] in the correct units. For a percent-of-height convention, enter 1.72. For a length convention in millimetres, enter 21 (or 10.5 if the half-height convention is required). For a volume convention, enter the volume of the parabolic section at full fill — which the user has computed as 78 L, but this is only correct if the firmware documentation explicitly states that A is a volume contribution in the chosen engineering units.
- Enter Dimension L [22] in the matching units. 98.28 %, 1199 mm, or 1034 L (the volume of the straight cylindrical section at full fill, computed as 1112 L minus the 78 L parabolic contribution), as appropriate for the firmware convention.
- Write parameters to the device. Issue a HART write to each parameter and verify the echo from the device matches the entered value. SIMATIC PDM will display a confirmation dialog; the local LCD will show the new value within two seconds.
- Apply damping and trim as required. Default damping is 0 s. For turbulent filling, 0.5–2.0 s is typical. Do a sensor trim only after the geometry is correct; trimming on bad geometry masks the underlying error.
Verification Procedure
With the geometry corrected, the transmitter must be verified at multiple fill levels. A four-point check at 0 %, 25 %, 50 %, 75 %, and 100 % is the field standard.
- Drain the vessel completely. Confirm the 4 mA output reads 4.000 ± 0.002 mA and the HART primary variable reads 0 L ± 1 L. If it does not, perform a zero trim (lower sensor trim) at this point with the diaphragm vented to atmosphere and the vessel empty.
- Fill to 25 % of the calibrated volume. For 1112 L full scale, target 278 L. The HART PV should read 278 ± 14 L (5 %). The 4–20 mA output should be 8.000 ± 0.080 mA. If the reading is high, A has been entered too small; if low, A has been entered too large. Adjust A in 0.1 % increments and repeat.
- Fill to 50 %. Target 556 L. PV should be 556 ± 14 L. This point confirms linearity of the cylindrical section; any error here points to the L dimension, not A.
- Fill to 75 %. Target 834 L. Same tolerance.
- Fill to 100 %. Target 1112 L. PV should match within 1 %. If only this point is off, the URV [3] is mis-set or the actual fill height is less than 1.22 m. Do not adjust Dimension A to fix the 100 % point; adjust the URV.
- Document the as-found and as-left values. Record the original (incorrect) A value, the corrected value, the four test-point readings, the HART firmware revision, and the date/time. File this with the loop documentation for the next calibration cycle.
Diagnostic Matrix for Common Failure Modes
| Symptom | Likely Cause | Corrective Action |
|---|---|---|
| Output flat at 4 mA below ~25 % fill, then linear | A [21] entered as ~25 % instead of geometrically correct value | Re-enter A in correct unit (typically 1.72 % for a 21 mm parabola in a 1220 mm vessel) |
| Output linear but offset low by a constant percentage | L [22] mis-set, A correct | Re-derive L from straight-section geometry |
| Output linear but offset high by a constant percentage | URV [3] under-set, geometry correct | Re-verify hydrostatic column at full fill, re-enter URV |
| Output reads 3.8–4.2 mA with empty vessel (slightly negative) | Diaphragm elevation above true low point; no air pocket at diaphragm | Document dead band; either accept it or relocate the nozzle closer to the apex |
| Output reads 3.6 mA with empty vessel and slow rise on fill | Air pocket trapped at diaphragm, or impulse line not fully filled | Vent the impulse line; verify process connection is flooded; check for vertical loops in the impulse line |
| Output noisy even with steady level | Damping set to 0, turbulent fill, or vibration | Increase damping to 0.5–2.0 s; verify mechanical isolation from pumps |
| Output tracks correctly at low fill but saturates below 100 % | URV [3] too low for actual head at full fill; or A [21] over-sized | Verify column height at overflow; re-derive URV; re-verify A |
| Output drifts over hours/days with steady level | Temperature effects on fluid density, or diaphragm fouling | Check process temperature stability; clean diaphragm if coated; re-zero with vessel empty |
| HART communication intermittent | Loop resistance > 600 Ω, or shield/grounding issue | Measure loop resistance; verify shield is grounded at one end only; check for multi-ground loops |
Loop and Electrical Checks
The SITRANS P320 is a two-wire 4–20 mA + HART device. Verify the loop is compliant before debugging the geometry:
- Supply voltage at the transmitter terminals: 10.5–45 V DC, depending on HART communicator load and certification. At 24 V nominal, 4–20 mA + HART, the typical operating voltage at the terminals is 12–30 V.
- Total loop resistance (wiring plus receiver input) must be in the HART-compliant range. For HART communication, the transmitter sees the receiver as a 250 Ω load; total loop resistance must be between 250 Ω and 600 Ω for reliable communication.
- Shield the cable and ground the shield at one point only (typically the cabinet end). Multi-grounded shields create ground loops that inject noise on the 4–20 mA signal.
- Do not run the signal cable in the same conduit as VFD output cables or other high-noise sources. If parallel runs are unavoidable, maintain 300 mm separation and cross at 90°.
Process-Side Considerations
Several process conditions can shift the indicated volume even when the geometry is correct:
- Density variation. The P320 has no density input. If the process fluid density changes with temperature or composition, the indicated volume will drift. For water between 4 °C and 40 °C, density varies by less than 0.5 %, which is usually negligible. For hydrocarbons with significant thermal expansion, a density correction must be applied externally.
- Vapor pressure and condensation. In a closed tank, condensation can accumulate in the impulse line and produce a constant head offset. A purge cycle or a condensate pot is required for steam or hot-vapor service.
- Aeration and foaming. Hydrostatic pressure measures the mass of the column above the diaphragm, regardless of phase. Foaming liquids will read heavier than the actual liquid volume. Antifoam injection at the dip pipe or a different measurement technology (guided wave radar) is required for foaming services.
- Agitator and pump-induced turbulence. Local pressure fluctuations at the diaphragm add noise to the 4–20 mA signal. Damping of 1–2 s smooths the output without lagging the actual level change by more than the operator can tolerate.
Scaling Check at a Single Test Point
If a full 0/25/50/75/100 % sweep is impractical, a single-point check at 50 % fill is the minimum verification. With geometry corrected and the tank at exactly half volume (556 L of 1112 L), the HART PV must read 556 L within 1 %. The 4 mA output must read 12.000 mA within 0.060 mA. If both are on target, the four-point linearity is very likely correct. A single-point failure usually points to a bad A or L value; a single-point pass with the other three points off would point to a URV/LRV mismatch and warrants a re-derivation of the hydrostatic column.
When to Contact Siemens Support
Escalate to Siemens Industry Online Support (support.industry.siemens.com) if any of the following are true after the above procedure has been executed:
- The 4-point verification still fails by more than 1 % after the geometry has been independently re-derived from the physical vessel drawing.
- The HART communicator cannot identify the device or reports a firmware revision that is not in the manual revision history.
- The diagnostic codes reported by the device (in SIMATIC PDM or on the LCD) include any of the following patterns: S−prefix codes indicating sensor failure, C−prefix codes indicating calibration invalid, or A−prefix codes indicating electronics failure. These are out-of-tolerance conditions that the geometry change cannot fix.
- The firmware revision is older than the current production release and exhibits the same behaviour after geometry correction; a firmware update may be required.
Long-Term Maintenance Recommendations
- Calibration interval. 12 months is typical for hydrostatic level on a non-fouling, non-corrosive service. Quarterly checks are recommended for slurries, paper stock, or any service where diaphragm coating is expected.
- Diaphragm inspection. At each calibration, isolate the transmitter, vent both sides, and inspect the diaphragm through the process connection. Mineral deposits, polymer buildup, or pitting require cleaning or replacement.
- Gasket and O-ring replacement. Replace the process connection gasket at every second calibration interval, or per the site maintenance procedure. A leaking gasket introduces a variable head loss that biases the reading.
- HART device list. Keep a current HART device list with tag, descriptor, message, firmware revision, and date of last calibration. This simplifies both scheduled maintenance and unscheduled troubleshooting.
- Loop validation. At each calibration, perform a 4–20 mA loop test from the device (in SIMATIC PDM or via HART command 40). Confirm that 4.000 mA, 12.000 mA, and 20.000 mA are produced on demand and read correctly at the controller or DCS.
What is the correct value of Vessel Dimension A for a 1.22 m parabolic-bottom vessel with a 21 mm parabolic section?
If A is expressed as a percentage of total vessel height, A = (21 / 1220) · 100 = 1.72 %. If A is expressed as a length in mm, A = 21 mm (or 10.5 mm if the firmware requires the half-height of the parabola as the input convention). Read the unit symbol next to the value on the device LCD or in SIMATIC PDM to determine which convention is in force for the installed firmware revision.
Why does the SITRANS P320 read zero until the tank is 25 % full when configured for PARAB?
The most common cause is that Vessel Dimension A [21] has been entered in the wrong engineering units. A value of ~25 % entered where ~1.7 % is correct shifts the volume model so the lower quartile of physical fill falls into the calculation dead zone. Re-derive A from the geometry (parabola height / total vessel height) and re-enter the value in the firmware's native unit for the parameter.
Where should the SITRANS P320 be mounted on a vertical vessel with a parabolic bottom?
The transmitter should be mounted on the bottom head, as close to the apex of the parabolic section as the nozzle allows. The diaphragm must be the lowest wetted point of the measurement column. Any vertical offset between the diaphragm and the true low point of the tank creates an irreducible dead band that must be documented and accepted, or eliminated by relocating the nozzle.
What is the difference between PARAB and PARAE in the Application parameter [5]?
PARAB is for a vertical cylinder with a single parabolic (domed) bottom; PARAE is for a horizontal cylinder with parabolic ends on both sides. Both modes use the same internal arithmetic of L + (2 · A), but the physical interpretation of the L and A dimensions differs. Select PARAB for vertical tanks with one dished head, and PARAE for horizontal tanks with two dished heads.
What is the URV (Upper Range Value) for a 1.22 m water column in a SITRANS P320?
Using P = ρ · g · h with ρ = 1000 kg/m³, g = 9.81 m/s², and h = 1.22 m, the full-scale pressure is 11 968 Pa, or 119.68 mbar. Round up to 120 mbar to give 0.3 % headroom for instrument tolerance and to ensure 20.000 mA output at the actual physical full point. For non-water services, recalculate with the actual fluid density.
Can the SITRANS P320 compensate for fluid density changes?
No. The SITRANS P320 measures gauge pressure and converts it to volume using fixed geometry parameters; it has no density input. If the process fluid density varies by more than 1 % over the operating temperature or composition range, the indicated volume will drift. Use an external density compensation (from a densitometer or a temperature-corrected density table) and re-scale the volume output, or select a different measurement technology such as guided wave radar for density-variable services.