SWAS Steam and Water Analysis System Commissioning: Yokogawa CA800 and ABB AWT420 Field Guide

SWAS Steam and Water Analysis System Commissioning: Yokogawa CA800 and ABB AWT420 Field Guide

Why SWAS Commissioning Failures Are Expensive

A SWAS panel looks simple: a few analyzers, some sample lines, and a drain. In practice, commissioning errors produce chemistry data that operators trust but should not. High conductivity readings masked by air ingress, pH offsets caused by contaminated reference electrodes, and dissolved oxygen alarms disabled during startup because “they always trip” — these patterns cause hundreds of millions in turbine damage every year.

Yokogawa CA800 and ABB AWT420 are widely deployed in utility and industrial boilers. Both require specific sample conditioning before any calibration begins. Calibrating an analyzer on unconditioned sample produces a calibration record that is useless from day one. EPRI Technical Report TR-1003138 defines the chemistry alarm setpoints for once-through and drum boilers. Your SWAS commissioning must deliver readings accurate enough to trigger those alarms at the correct process values.

Sample Line Pre-Conditioning: The Step Most Engineers Skip

Before powering on any analyzer, the sample lines need 72 hours of continuous flushing at full flow. This removes mill scale, welding residue, and atmospheric contamination that accumulates during construction. Failure to pre-condition produces pH readings 0.5–1.0 units above actual process values for the first two weeks of operation.

Set sample pressure at the SWAS panel inlet to between 0.5 and 1.0 bar gauge. Sample temperature at the analyzer inlet must be below 40°C for both Yokogawa CA800 and ABB AWT420. Verify this with a calibrated contact thermometer before connecting analyzer tubing. Flow rate through the conductivity cell on the CA800 should be 100–200 mL/min. For the ABB AWT420 pH cell, flow must be 50–150 mL/min at the reference junction. Flush until the online conductivity reading stabilizes within ±0.05 µS/cm over a 30-minute window. Document the flush duration and stable reading as the pre-commissioning baseline.

Yokogawa CA800 Conductivity Calibration Procedure

The CA800 uses an electrodeless toroidal conductivity sensor. Use NIST-traceable 84 µS/cm or 1413 µS/cm KCl standard solutions, depending on your expected process range.

  • Step 1: Open the CA800 front panel and navigate to Calibration Menu → Conductivity Calibration → One-Point Offset.
  • Step 2: Remove the sensor from the sample cell. Rinse with deionized water (resistivity >1 MΩ·cm). Dry with lint-free cloth.
  • Step 3: Immerse the sensor fully in the 84 µS/cm standard solution. Wait 5 minutes for temperature equilibration.
  • Step 4: Record the displayed value. Acceptable tolerance is ±0.5 µS/cm. If outside this range, adjust the cell constant in the CA800 configuration menu (typical range 0.095–0.105 cm⁻¹).
  • Step 5: Rinse the sensor, reinstall in the sample cell, and restore sample flow. Verify the reading stabilizes within 2 minutes to within ±0.2 µS/cm of the pre-rinse baseline.

For cation conductivity measurement (after a hydrogen cation exchange column), the expected process value in all-volatile treatment (AVT) mode is below 0.2 µS/cm. Calibrate the CA800 with a 0.1 µS/cm standard in this range. Do not use the 84 µS/cm standard for cation conductivity — the measurement uncertainty at low concentrations becomes unacceptably large.

ABB AWT420 pH and Dissolved Oxygen Calibration

The ABB AWT420 is a multi-parameter transmitter supporting pH, ORP, dissolved oxygen, and conductivity on separate sensor inputs. For pH calibration, use a two-point procedure with NIST pH 4.01 and pH 7.00 buffers. Access the AWT420 calibration wizard via the front keypad: Menu → Calibration → pH → Two-Point. Acceptable slope range is 53–62 mV/pH at 25°C. A slope below 50 mV/pH indicates a contaminated or depleted reference junction — replace the electrode before proceeding.

For dissolved oxygen (DO) calibration, use air saturation at known temperature. The AWT420 DO sensor uses a polarographic membrane. Calibrate in air at 100% saturation: Menu → Calibration → DO → Air Calibration. Enter the current barometric pressure (typical sea-level value 101.325 kPa). Boiler feedwater DO alarm setpoints follow EPRI guidelines: AVT(O) oxygen target is 30–150 ppb. Configure AWT420 relay R1 to alarm at 10 ppb low and 200 ppb high for AVT mode. Use 4–20 mA output scaled 0–500 ppb for historian trending. Do not disable the low DO alarm during startup — oxygen scavenger dosing faults often appear first as a low DO trip.

Common SWAS Fault Patterns and Diagnostic Steps

  • Fault 1 — CA800 reading drifts high overnight: Air ingress through a loose fitting upstream of the sensor. Check all compression fittings at the sample cooler outlet. Retorque to manufacturer specification (typically 1.5 N·m for 6 mm Swagelok). Verify sample pressure remains above 0.3 bar gauge at all flow conditions.
  • Fault 2 — ABB AWT420 pH reading oscillates ±0.3 pH units: Reference junction pressure instability from excessive sample flow. Reduce sample flow to 80 mL/min and observe for 15 minutes. If oscillation stops, install a back-pressure regulator at the cell outlet set to 0.2 bar.
  • Fault 3 — DO reading shows 8–9 mg/L (air saturation) in feedwater: Membrane fouling or cracking. Replace the polarographic membrane and recalibrate. Inspect the sample line for air entrainment at the pump suction connection.
  • Fault 4 — Cation conductivity reading shows negative value: Depleted or exhausted cation exchange resin column. Measure the outlet pH from the cation column. If above 7.0, the resin is exhausted and must be replaced. Resin service life at 100 L/day sample flow is typically 6–12 months.

Document each fault with the analyzer tag, fault description, root cause, and corrective action in your CMMS. SWAS fault history is a key indicator of overall water chemistry program health and appears in EPRI BenchmarkingPlus assessments.

Conclusion and Action Advice

SWAS commissioning requires the same rigor as any safety-critical instrument calibration. Yokogawa CA800 and ABB AWT420 both deliver accurate measurements when installed and calibrated correctly. The difference between a functional SWAS and a decorative one is 72 hours of sample line pre-conditioning, traceable buffer calibration, and a fault response procedure that operators actually use.

Review your current SWAS documentation this week. If you cannot find a calibration record with as-found and as-left values for each analyzer, your data is not auditable. Implement the calibration sequences above and tie each record to a CMMS work order. EPRI chemistry compliance begins with reliable instruments — and reliable instruments begin with a disciplined commissioning process.

Author: Liu Jianguo is an industrial automation engineer with over 10 years of experience in PLC, DCS, and control systems.

Show All
Blog posts
Show All
Single Acting vs Double Acting Actuators: Selection Criteria, Fail-Safe Design, and Field Fault Diagnosis

Single Acting vs Double Acting Actuators: Selection Criteria, Fail-Safe Design, and Field Fault Diagnosis

Engineering-level comparison of single acting and double acting actuators with spring force calculations, solenoid wiring differences, and fault isolation procedures for Woodward ProAct and Allen-Bradley valve control applications.
SWAS Steam and Water Analysis System Commissioning: Yokogawa CA800 and ABB AWT420 Field Guide

SWAS Steam and Water Analysis System Commissioning: Yokogawa CA800 and ABB AWT420 Field Guide

Sample line pre-conditioning, multi-parameter calibration, fault isolation, and EPRI-aligned alarm setpoints for boiler chemistry protection using Yokogawa CA800 and ABB AWT420 analyzers.
IEC 61511 Safety Bypass and Override Management: HIMA HIMatrix F60 and Triconex T3000 Practical Guide

IEC 61511 Safety Bypass and Override Management: HIMA HIMatrix F60 and Triconex T3000 Practical Guide

IEC 61511 Clause 11.9 makes bypass management a mandatory lifecycle element. This guide covers the full workflow from bypass classification to closure for HIMA HIMatrix F60 and Triconex T3000, including hardware inhibit procedures, FORCE_DO configuration, PFDavg impact calculation, and a five-step audit trail process.