Quick Answer
Dissolved oxygen is monitored in boiler feedwater to confirm that the steam-water cycle is operating within its intended chemistry regime.
The correct dissolved-oxygen condition is not the same for every plant. Excess oxygen can contribute to corrosion risk under some conditions, particularly when ionic contamination is present. But extremely low oxygen combined with reducing chemistry can also promote iron transport and flow-accelerated corrosion in susceptible systems. For that reason, dissolved oxygen should be controlled against the plant’s specified cycle chemistry rather than driven blindly towards zero. [1]
In power-cycle applications, the challenge is not merely choosing a DO sensor. Reliable measurement depends on the sample point, trace measurement range, sample cooling, sample-line design, flow, possible oxygen ingress, analyzer technology, calibration and maintenance.
Why Is Dissolved Oxygen Measured in Boiler Feedwater?
Water returning through the condensate and feedwater system comes into contact with metallic equipment over a large surface area. The chemical environment around that metal strongly influences corrosion behaviour and the formation of protective oxide layers.
Dissolved oxygen is therefore one of the parameters used to verify that the feedwater treatment is producing the intended environment.
IAPWS identifies two reasons for controlling and monitoring DO. At one end, elevated oxygen combined with ionic contaminants—particularly chlorides—can increase corrosion risk. At the other, very low oxygen in conjunction with a reducing agent can contribute to enhanced iron transport and flow-accelerated corrosion where all-volatile treatment is used. [1]
This means a DO analyzer is not simply a corrosion alarm. It is an instrument for checking whether the actual oxidation-reduction environment is consistent with the selected chemistry programme.
Why “Lower Dissolved Oxygen Is Always Better” Is the Wrong Rule
Power-cycle chemistry may use different treatment philosophies.
| Treatment concept | General oxygen philosophy | Interpretation of a DO reading |
|---|---|---|
| AVT(R) — reducing all-volatile treatment | Reducing conditions are maintained and oxygen is generally controlled to a low level | An unexpectedly high result can indicate loss of oxygen control, but the complete chemistry programme must be considered |
| AVT(O) — oxidizing all-volatile treatment | No reducing agent is used in suitable all-ferrous systems; mildly oxidizing conditions are maintained | Some oxygen is compatible with the intended passive condition; “zero” is not automatically the goal |
| OT — oxygenated treatment | Oxygen is intentionally introduced and controlled in suitable systems | A low reading can itself indicate failure to achieve the selected chemistry regime |
IAPWS publishes plant- and treatment-specific targets rather than one universal DO limit. Its volatile-treatment guidance, for example, distinguishes AVT(R), AVT(O) and oxygenated treatment and ties their suitability to plant metallurgy and cycle configuration. [2]
Engineering Note
A DO value should therefore be interpreted only after answering:
What feedwater treatment is this unit actually operating on, and what is the approved target for that plant and operating condition?
An analyzer reading that is undesirable for one chemistry regime may be intentional in another.
Where Should Dissolved Oxygen Be Measured?
The correct locations depend on plant design and the objective of the measurement.
For conventional plants, IAPWS includes dissolved-oxygen monitoring at the economizer inlet among key feedwater instrumentation because it confirms satisfactory oxygen control within the plant’s target regime. In HRSG applications, the analogous guidance refers to locations downstream of the feed pumps. Where oxygen dosing is practised, confirmation at the economizer inlet becomes particularly important. [1]
DO may also be measured upstream and downstream of the deaerator when the objective is to verify deaerator performance. IAPWS treats these as potentially useful additional measurements rather than prescribing one identical arrangement for every plant. [1]
A practical way to think about the measurement points is:
| Measurement point | What it can help establish |
|---|---|
| Condensate system | Whether oxygen is entering or changing in the condensate/feedwater path |
| Upstream of deaerator | Oxygen level entering the deaeration stage |
| Downstream of deaerator | Whether the deaerator is producing the intended condition |
| Economizer inlet / final feedwater | Whether the water entering the boiler side is within the required chemistry regime |
| Selected diagnostic points | Used when investigating air ingress, chemistry changes or treatment performance |
ABB likewise identifies condensate DO monitoring as useful for detecting inefficient aeration or air leakage and uses low-level DO measurement in boiler-feedwater applications. [3]
The purpose of the measurement should therefore be defined before deciding whether one analyzer, multiple dedicated analyzers or a switched sampling arrangement is appropriate.
Why Boiler-Feedwater DO Is a Trace Measurement Problem
A general dissolved-oxygen analyzer may be designed for water treatment or aeration where concentrations are conveniently expressed in mg/L.
Power-cycle feedwater can require measurement much further down into the µg/kg, or approximately ppb, region.
IAPWS notes that calibration and maintenance procedures must be especially careful when accuracy is required in the low-µg/kg range. Commercial instruments designed specifically for power-cycle chemistry likewise provide trace-oxygen capability; for example, SWAN markets dedicated analyzers for low-ppb dissolved oxygen in high-purity power-cycle water, while METTLER TOLEDO offers both optical and polarographic sensors intended for power-plant cycle chemistry. [1][4][5]
Engineering Note
Measurement range and resolution matter.
An instrument being labelled a “dissolved oxygen analyzer” does not automatically make it suitable for boiler-feedwater trace measurement. The analyzer must be capable of resolving the actual chemistry range with the required accuracy and response.
How Is Dissolved Oxygen Measured Online?
IAPWS’s 2024 guidance identifies two main technologies for online power-cycle DO measurement. [1]
1. Membrane / amperometric measurement
Dissolved oxygen passes through a permeable membrane and participates in an electrochemical reaction at the sensor. The resulting electrical current is related to the oxygen concentration.
This family includes polarographic/amperometric designs widely used for trace oxygen measurement.
2. Luminescence / optical measurement
An optical sensor uses an oxygen-sensitive luminescent material. The material is excited optically, and the presence of oxygen influences its luminescence behaviour. By evaluating parameters such as decay time, the instrument determines oxygen concentration.
Importantly, IAPWS states that both technologies can provide highly reliable and accurate readings in typical power-system applications when the measurement system and procedures are suitable. [1]
The selection decision should therefore not be reduced to “electrochemical versus optical” in isolation. Required range, response behaviour, calibration, maintenance, sample arrangement and the specific analyzer design all matter.
Why Sample Conditioning Can Matter as Much as the Analyzer
A technically capable DO analyzer can still report the wrong process condition if the sample reaching it has changed between the process take-off point and the sensor.
Safety Note
This is particularly important in steam-water analysis because the original process can be at high temperature and pressure, while the analyzer normally receives a controlled liquid sample through a sampling and conditioning system.
TIPL’s SWAS solution is built around continuous monitoring of steam-water chemistry at feedwater, condensate, deaerator, drum, steam and other critical points. [6]
For dissolved oxygen specifically, IAPWS warns of several sample-line effects.
Residual reducing agents or other reactive species may consume oxygen while the sample travels to the analyzer, creating an artificially low reading. The sample-line wall itself may also exert oxygen demand under changing redox conditions. [1]
IAPWS therefore recommends measures including:
- cooling the sample as quickly as practical;
- minimizing sample-line length;
- maintaining proper sample flow; and
- minimizing fittings upstream of the DO analyzer to reduce opportunities for atmospheric oxygen ingress. [1]
Engineering Principle
The DO analyzer is measuring the sample that reaches its sensor—not the sample that originally left the process.
The sampling system must preserve the parameter well enough for those two conditions to represent each other.
What Can Make a Boiler-Feedwater DO Reading Misleading?
| Condition | Possible measurement effect | What to check |
|---|---|---|
| Air entering the sample line | Reading can become higher than the true process condition | Fittings, tubing, connections and sampling arrangement |
| Residual reducing chemistry reacting in the sample line | Reading can become lower before the sample reaches the sensor | Chemistry regime, transport time, cooler location and line length |
| Inadequate or changing sample flow | Slow or inconsistent response; poor representation of the current process | Flow through the sampling system and analyzer flow cell |
| Long sample line | Greater opportunity for reaction and response delay | Routing and sample-point-to-analyzer distance |
| Poor sample cooling | Can increase reaction during transport and create unsuitable analyzer conditions | Cooler location and sample-condition design |
| Wrong measurement range | Analyzer may lack useful resolution in the required trace region | Published analyzer range, resolution and low-level performance |
| Calibration or maintenance problem | Offset, slow response or questionable low-level accuracy | Manufacturer’s calibration and maintenance procedure |
| Wrong chemistry target | A technically accurate measurement may be interpreted incorrectly | Approved plant chemistry programme and operating state |
Not every abnormal DO value is therefore an analyzer fault.
The first question is whether the instrument is measuring the sample correctly. The second is whether the sample still represents the process. The third is whether the engineer is comparing the value against the correct chemistry target.
How Should a High Dissolved-Oxygen Reading Be Interpreted?
A high reading should trigger investigation, not an automatic chemical response.
Possible areas to examine include:
- actual feedwater chemistry;
- condenser or system air ingress;
- deaerator operation where applicable;
- the current operating/load condition;
- sample-line leakage;
- calibration and analyzer condition; and
- whether the reading is being compared with the correct cycle-chemistry target.
For example, DO measurements upstream and downstream of a deaerator can help determine whether the deaeration stage is performing as expected. [1]
However, simply adding or increasing an oxygen scavenger is not a universal corrective action. Reducing-agent use itself depends on the selected cycle chemistry, metallurgy and plant configuration. IAPWS’s volatile-treatment guidance specifically distinguishes reducing and oxidizing programmes. [2]
Can a Very Low DO Reading Also Be a Problem?
Yes—depending on the chemistry regime.
A very low value may be completely appropriate for one plant. In another, especially where oxygenated treatment or an oxidizing all-volatile programme is intended, it can indicate that the required chemistry environment is not being maintained.
IAPWS also warns that very low oxygen combined with a reducing agent can create conditions associated with enhanced iron transport and flow-accelerated corrosion in susceptible feedwater systems. [1]
There is also a measurement question: an unexpectedly low result can arise because oxygen is being consumed inside the sample line before reaching the analyzer.
So a low number should not automatically be accepted as “good”.
Calibration and Maintenance at Trace Levels
Trace-level DO measurement places much more emphasis on measurement quality than a routine high-range water measurement.
IAPWS states that calibration and maintenance of both amperometric and luminescence systems require careful procedures when measurements are being made in the low-µg/kg range. [1]
The exact method and frequency should follow the analyzer manufacturer’s documentation and the plant’s quality procedure rather than a generic interval.
Useful checks include:
- verify analyzer diagnostics and sensor condition;
- verify calibration using the manufacturer’s procedure;
- inspect the sample flow;
- confirm sample temperature and conditioning;
- check the low-level response rather than only a convenient high-level calibration point where required by the analyzer design; and
- compare abnormal trends with other chemistry parameters rather than diagnosing DO in isolation.
A Practical Check Before Trusting a Boiler-Feedwater DO Value
- The correct target — Which cycle chemistry treatment is being operated?
- The correct sample point — What process condition is this location intended to represent?
- The correct analyzer range — Can the measuring system resolve the required trace level?
- The sampling system — Is the sample cooled promptly and transported with minimal opportunity for reaction or air ingress?
- The sample flow — Is the instrument receiving the required stable flow?
- Analyzer condition — Are calibration, sensor condition and diagnostics satisfactory?
- Supporting evidence — Do conductivity, pH, chemistry treatment, deaerator operation and other relevant plant data support the same conclusion?
A single DO number is most useful when it is interpreted as part of the steam-water chemistry system rather than as an isolated parameter.
Key Takeaway
Dissolved-oxygen measurement in boiler feedwater is fundamentally a cycle-chemistry control measurement.
The objective is not simply to obtain the lowest possible oxygen reading. The objective is to verify that the plant is operating within the dissolved-oxygen condition required by its selected chemistry regime.
Reliable measurement therefore requires four things to work together:
correct chemistry target → correct sampling point → representative conditioned sample → trace-capable analyzer.
If any one of these is wrong, the displayed value can lead to the wrong engineering conclusion.
Evaluating Dissolved Oxygen as Part of SWAS
In a steam-water cycle, dissolved oxygen should be evaluated alongside the required chemistry parameters and sampling locations rather than selected as a standalone water-quality instrument.
TIPL’s Steam and Water Analysis System (SWAS) covers continuous chemistry monitoring across feedwater, condensate, deaerator, drum, steam and other critical plant points. The required DO range, sampling arrangement, plant chemistry and analyzer technology should be established before the final instrument is selected. [6]
Technical References
- International Association for the Properties of Water and Steam (IAPWS) — TGD2-09(2024), Instrumentation for monitoring and control of cycle chemistry. Guidance on dissolved-oxygen measurement, low-µg/kg accuracy, sample-line effects and power-cycle monitoring locations.
- IAPWS — TGD3-10(2015), Volatile treatments for the steam-water circuits of fossil and combined cycle/HRSG power plants. Guidance distinguishing AVT(R), AVT(O) and oxygenated treatment and requiring plant-specific chemistry targets.
- ABB — Continuous Water Analysis for Power Generation. Supporting manufacturer guidance on dissolved-oxygen monitoring in condensate and boiler-feedwater applications.
- SWAN Analytical Instruments — AMI Oxytrace QED. Manufacturer documentation for continuous dissolved-oxygen measurement in high-purity water with ppb-range resolution.
- METTLER TOLEDO — High-Purity Water Polarographic Dissolved Oxygen Sensor. Example of instrumentation designed for power-plant cycle chemistry and demanding low-ppb applications.
- TIPL — Steam and Water Analysis System (SWAS). TIPL solution context for continuous chemistry monitoring across critical steam-water-cycle sampling points.
- TIPL — Power Industry Instrumentation Solutions. TIPL power-industry context for SWAS and dissolved-oxygen monitoring.
