Zero drift is among the most insidious flow metre faults: the instrument appears to be working, but at zero flow (no movement) it outputs a non-zero signal. A metre outputting 0.2 mA when flow is zero, and 4.2 mA at actual zero flow, is lying about measurement. Over time, zero flow errors compound into large cumulative measurement errors. Understanding zero drift—its causes, detection, and prevention—is essential for custody transfer applications and high-accuracy energy balances.
Defining Zero Offset and Zero Drift
Zero Offset (Static)
The difference between the output signal at true zero flow and the expected zero signal (usually 4 mA for 4–20 mA transmitters). Example: true zero flow produces 4.15 mA instead of 4.00 mA = 0.15 mA offset. Acceptable limit: typically ±0.1–0.2 mA depending on accuracy class.
Zero Drift (Dynamic)
Zero offset that changes over time. Example: at installation, zero offset is 0.05 mA; after 6 months, it is 0.40 mA. The metre's zero point is drifting. If left unchecked, zero offset will eventually exceed the acceptable limit and compromise measurement accuracy.
Why Zero Drift Occurs: Root Causes
Electromagnetic Metres
Primary cause: Electrode fouling
The electrodes measure signal strength in the conductive fluid. Accumulation of slime, mineral scaling, or biofilm on electrode surfaces reduces signal amplitude, causing the electronics to increase gain to compensate. This gain increase is mistaken for zero drift (the "zero" reference creeps upward). Over weeks to months, zero offset increases visibly.
Secondary causes: Temperature drift in electronics (0.1–0.3 mA per 10 °C in older designs); cable insulation degradation in corrosive environments (increased leakage current).
Coriolis Metres
Primary cause: Tube zero-point calibration drift
Coriolis metres measure phase difference between inlet and outlet tube vibration. At zero flow, phase difference should be exactly zero. Electronic temperature-compensation circuits adjust the zero reference based on temperature. If temperature compensation is inaccurate or the sensor itself has aged (piezo ceramic degradation), zero reference shifts. Typical drift: 0.05–0.15% per year in bench-top units; much lower (<0.05% per year) in modern transmitters with digital temperature compensation.
Secondary cause: Mechanical resonance from external vibration (machinery, pump ripple) can be mistaken for flow signal, causing artificial zero offset (usually spiky, not gradual).
Vortex Metres
Primary cause: Shedding pin wear or deposit accumulation
The shedding pin oscillates at a frequency proportional to flow. At zero flow, oscillation should be absent; the signal should be quiet. Wear of the pin or accumulation of deposits can cause small oscillations even at zero flow, generating a false signal (phantom zero offset). Less common than EM or Coriolis drift, but possible in dirty applications.
Turbine Metres
Primary cause: Bearing wear or rotor drag
Turbine rotors are free-spinning and should coast to a stop with no output when flow stops. Bearing wear, corrosion on the rotor, or viscosity changes can introduce a frictional "creep" signal. The magnetic pickup coil may sense rotor vibration or slow movement, producing a phantom zero-flow signal. Typical offset: 0.02–0.1 Hz (equivalent to ~0.1–0.5% of full-scale flow).
Technologies Most Susceptible to Zero Drift
High risk (frequent zero offset issues):
- Electromagnetic metres in slurry or high-particulate applications (monthly zero checks recommended)
- Turbine metres in corrosive or high-viscosity fluids (biannual zero checks)
Medium risk (occasional drift):
- Coriolis metres in very high-temperature applications (>100 °C; thermal drift possible)
- Vortex metres in fouling-prone applications (biofilm, mineral scaling)
Low risk (rare zero drift):
- Modern Coriolis metres with digital temperature compensation in normal conditions
- Ultrasonic clamp-on metres (no internal wetted parts; nearly immune to zero drift)
Detecting Zero Offset
Method 1: Block-and-Measure Test (Field)
Isolate the metre by closing inlet and outlet block valves (creating zero flow). Record the signal output (4–20 mA transmitter should read 4.00 mA ±0.2 mA tolerance). If output is outside tolerance, zero offset is present. Repeat test monthly for critical applications (custody transfer, energy balance).
Equipment needed: Multimeter (cost: GBP 50–200)
Method 2: Loop Current Measurement
For 4–20 mA transmitters, record the current at the receiver/PLC input. At zero flow, current should be exactly 4 mA (or within ±0.2 mA). Trend the value over months to detect drift. Many modern process control systems log this data automatically.
Method 3: HART or Modbus Diagnostics
If the metre supports HART protocol (electronic hand-held communicator) or Modbus (Ethernet), connect and read the "zero output" or "percent of range" diagnostic. Many smart metres store the zero calibration value internally; comparison against a baseline reading reveals drift.
Equipment needed: HART communicator (GBP 800–2,000) or Modbus gateway with laptop
Method 4: Trend Analysis
For high-accuracy applications, maintain a trend log of zero-flow readings collected monthly or quarterly. Plot the zero signal over time. A slope >0.02 mA per month indicates drift; take corrective action before offset exceeds tolerance.
Auto-Zero and Zero Adjustment Features
Manual Zero Adjustment
Many transmitters include a manual zero-adjustment potentiometer or button. Procedure: isolate metre (zero flow), read current output, adjust trim pot until output reads exactly 4.00 mA. Cost: free (on-site adjustment). Limitation: manual adjustment is temporary; drift will resume unless root cause (fouling, temperature) is addressed.
Automatic Zero Tracking (Soft-Ware Based)
Modern Coriolis and EM metres include auto-zero algorithms. At power-up or on demand, the transmitter measures zero flow and sets the zero reference internally. Common implementation: zero set on power-up + optional periodic zero check (e.g., daily at midnight if flow is detected to be zero). Benefit: drift is corrected automatically without manual intervention. Limitation: if fouling is the root cause, auto-zero may mask the problem until electrode cleaning is needed.
Drift Detection Algorithm
Some smart metres monitor zero offset continuously. If drift exceeds a threshold (e.g., 0.3 mA), the metre logs an alarm in HART/Modbus diagnostics. Technician is alerted to perform maintenance (e.g., electrode cleaning for EM metres). Example: Endress+Hauser Promass 80F Coriolis metres include "zero offset monitoring" which flags >0.15% drift.
Temperature Compensation (Automatic)
Coriolis and some EM metres include temperature sensors. Internal electronics adjust zero reference based on temperature. This reduces zero drift from thermal effects significantly (typical improvement: 10x reduction in drift rate from temperature-caused offset).
Prevention Strategies
1. Select Appropriate Technology
For applications with high fouling risk, prefer technologies less susceptible to zero drift: Coriolis (excellent zero stability) > ultrasonic (immune) > vortex > turbine > EM in problematic fluids.
2. Install Upstream Filtration or Conditioning
For EM metres: strainer (100–200 µm) upstream reduces electrode fouling. Cost: GBP 200–800. Benefit: extends electrode-cleaning intervals from 6 months to 12–24 months.
For turbine metres in viscous fluids: heater upstream to reduce drag. Cost: GBP 500–2,000.
3. Enable Auto-Zero on Power-Up
Configure transmitter to execute auto-zero calibration immediately after power-up and, if possible, daily at a known zero-flow time (e.g., midnight when plant is idle). Cost: zero (firmware feature). Benefit: automatic correction of slow zero drift.
4. Establish Quarterly Zero Checks
Use block-and-measure test to verify zero offset remains within specification. If offset exceeds limits, take corrective action (electrode cleaning, bearing inspection, or factory recalibration). Cost: 30 minutes per metre. Benefit: early detection before measurement accuracy is compromised.
5. Monitor Trend Data
For critical applications, plot zero-flow signal vs time. A linear trend indicates systematic drift requiring action; a step-change suggests sudden electrical or mechanical fault. Trend analysis is particularly valuable for identifying fouling in EM metres (zero offset increase rate indicates fouling rate).
6. Perform Planned Maintenance
Annual or biennial service by qualified technician includes zero calibration verification and electrode cleaning (EM) or rotor inspection (turbine). Cost: GBP 800–1,500 per visit. Benefit: prevents zero drift from becoming a chronic problem.
Real-World Example: Water Utility Custody Transfer
Scenario
A water utility operates an electromagnetic metre at the intake point. Responsibility to customer begins at this point; all flow beyond this metre is billable. After 18 months of operation, zero-offset trend shows creep from 0.08 mA to 0.35 mA (exceeding 0.2 mA specification).
Impact
At 4–20 mA range, 0.35 mA zero offset = 0.35 / 16 = 2.2% of full-scale flow. If full scale is 500 L/min, zero offset represents 11 L/min phantom flow. Over a month, 11 L/min × 60 min × 24 h × 30 days = 475 m³ false high reading = GBP 950+ billing error (at GBP 2 per m³).
Root Cause
Electrode fouling from algae biofilm in the intake water.
Solution
Clean electrodes (distilled water + soft brush). Cost: GBP 150 (30 min technician time). Zero offset drops to 0.06 mA (within specification).
Prevention
Install quarterly zero-offset trending. Install 150 µm strainer upstream (GBP 500 capital). Enable auto-zero algorithm on power-up (firmware). Result: zero-offset trend remains stable; no further action required for 3+ years.
Key Takeaways
- Zero drift is common in EM and turbine metres; less common but possible in Coriolis and vortex
- Root causes vary by technology: EM (electrode fouling), Coriolis (thermal compensation drift), turbine (bearing wear)
- Detection requires active monitoring: quarterly block-and-measure tests or monthly trend logging
- Auto-zero features help but don't solve fouling: manual electrode cleaning is often still needed
- Prevention is cost-effective: GBP 150–500 in preventive measures saves GBP 5,000+ in measurement errors or emergency repair