Precision instead of Washout Loss: Why Water-Soluble Gases Require Hot-Wet Analysis
In gas analysis, sampling with subsequent gas cooling is often considered standard. However, for water-soluble organic components, this “cold-dry method” reaches physical limits. Using formaldehyde and ethanol as examples, we demonstrate why hot-wet measurement (hot-wet analysis) is indispensable for valid results and what solutions Fresenius Umwelttechnik offers for this.
The Physico-Chemical Basis: Water Solubility and Washout
Highly polar molecules such as formaldehyde (HCHO) and ethanol (C₂H₅OH) preferentially dissolve in water. If the sample gas is cooled to a typical dew point of 4°C in a gas cooler, not only does water vapor condense, but the components to be measured are also washed out in the condensate. In the case of formaldehyde, this effect is further amplified by the almost complete hydration to methylene glycol:
HCHO + H₂O ⇌ CH₂(OH)₂
As a result, the effective dimensionless distribution coefficient between liquid and gas phase at 4°C is extremely large for formaldehyde (cliquid/cgas ≈ 3,000–5,000), and significantly smaller for ethanol, but still sufficient for significant washout losses.
Important physical principle: According to Henry’s Law, the distribution coefficient in the technically relevant concentration range is independent of the absolute concentration of the target component. The percentage washout loss is the same for every concentration under constant boundary conditions (temperature, humidity, flow rate, contact time). The absolute washed-out quantity increases proportionally with concentration.
Typical losses, depending on conditions, range from 30–60% and more for formaldehyde and 10–35% for ethanol (under constant conditions in each case). Gas cooling is fundamentally unsuitable for both components.
Practical Case 1: Formaldehyde Measurement in Asphalt Recycling Plants
During the reprocessing of asphalt, thermal processes generate various organic compounds, including formaldehyde. After activated carbon filtration, compliance with the emission limit value must be continuously monitored.
Important: EC sensors are only suitable for cool, dry ambient air. They cannot be used for hot, humid exhaust gases (hot-wet measurements) as they neither measure stably nor are sufficiently selective against accompanying gases and high humidity under these conditions.
Process Conditions and Requirements
The exhaust gas after the activated carbon filter has the following characteristic properties:
- Temperature: 40–50°C
- Humidity: 4–12 Vol.-% (physically possible saturation range at 40–50°C; at 40°C max. approx. 7.3 Vol.-%, at 50°C max. approx. 12.2 Vol.-%)
- Oxygen: 6–20.9 Vol.-%
- Carbon Dioxide: 5–25 Vol.-%
- Carbon Monoxide: 0–600 mg/Nm³
- Nitrogen Oxides: 50–300 mg/Nm³
- Naphthenes (CₙH₂ₙ): 200–500 mg/Nm³
- Required measuring range for formaldehyde: depends on the applicable limit value (for 5 mg/m³: measuring range 0–5 mg/m³; for 10 or 15 mg/m³: at least 0–20 mg/m³)
Why NDIR Technology is Not Suitable in This Case
Non-dispersive infrared spectroscopy (NDIR) is a proven technology for many gas components. However, for formaldehyde in this case, three fundamental limitations arise:
Insufficient detection limit: Typical NDIR devices achieve a detection limit of approx. 2 mg/m³ for formaldehyde. For a limit value of 5 mg/m³, the detection limit must be ≤ 1.0 mg/m³ according to the basic metrological rule (NDG ≤ limit value/5) – this requirement cannot be met with NDIR.
Cross-sensitivity to hydrocarbons: The naphthenes in the process gas can influence the NDIR formaldehyde signal for two reasons:
- Secondary transmissions (ghost transmissions) of the NDIR filters: Residual transmissions of typically 0.01–0.1% allow accompanying gases to contribute to the measurement signal.
- Overlap of the flanks of the absorption ranges: The flanks of the IR absorption bands of hydrocarbons and formaldehyde overlap, leading to a positive deviation of the formaldehyde signal.
Both effects can be reduced by an additional HC compensation channel, but they do not improve the detection limit.
Washout losses with gas cooling: Under the present humidity conditions (4–12 Vol.-%), washout losses of over 50% are realistic. Gas cooling is fundamentally unsuitable.
The advantage of the NDIR technology of the GA315 is that it is by far the most cost-effective method for measuring formaldehyde.
Recommended Measurement Technologies for Formaldehyde
- FTIR Spectroscopy (preferred solution): Detection limits 0.1–0.3 mg/m³, very high selectivity through multivariate analysis, hot-wet operation at approx. 180°C measuring cell temperature.
- CRDS Technology: Laser spectroscopy with extremely long optical path lengths, detection limits in the sub-mg range, very high selectivity.
- GA315 with HC compensation (theoretically possible): When applying the 15 mg/m³ limit value, the GA315 with implemented naphthene compensation could be used. For the 5 mg/m³ limit value, the detection limit of approx. 2 mg/m³ is not sufficient.
Practical Case 2: Ethanol Measurement in Yeast Production
In yeast fermentation processes, the ethanol concentration in the exhaust gas is an important process parameter for controlling fermentation. Measurement typically occurs in the 0-200 or 0-1000 ppm range. Ethanol is produced as a metabolic product of yeast and is continuously discharged with the fermentation gas.
Process Conditions in Fermentation
The fermentation gas is characterized by:
- Temperature: 30-45°C (water vapor saturated)
- Humidity: saturated; at 30°C max. approx. 4.2 Vol.-%, at 45°C max. approx. 9.6 Vol.-%
- Ethanol: 10-1000 ppm
- Dust content: <5 mg/m³
- Pressure: almost atmospheric (0-10 mm H₂O)
Quantification of Washout Losses
Ethanol is highly water-soluble. The percentage washout loss is independent of the ethanol concentration under constant conditions – this follows directly from Henry’s Law. However, losses increase with increasing humidity:
| Humidity (Process Gas) | Condensate Volume (approx.) | Ethanol Washout Loss (approx.) |
| 4.2 Vol.-% (30°C sat.) | low | 10–15% |
| 7.3 Vol.-% (40°C sat.) | medium | 15–25% |
| 9.6 Vol.-% (45°C sat.) | elevated | 25–35% |
These losses are not constant but depend on the ethanol concentration, the amount of condensate, and the residence time in the gas cooler. Reliable process control is therefore not possible.
Technical Solution Variant from Fresenius Umwelttechnik
The GA320 from Fresenius Umwelttechnik is the recommended solution for ethanol measurement:
- Multiple sample gas inputs via heated solenoid valves (70°C)
- Heated measuring cells: up to 60°C
- Internal sample gas paths made of stainless steel are consistently heated along with the input block and measuring cell – no cold zone in the sample gas path
- Integrated humidity compensation: H₂O compensation channel continuously corrects the cross-sensitivity of the ethanol signal
- Modular concept: Multi-component measurement possible
- Measurement principle: Exclusively Two-Pressure Reference Method (see Section 5)
In contrast, a conventional NDIR device with an upstream gas cooler would fundamentally work, but with the described systematic errors of 10–35%, depending on the process conditions. This solution is unsuitable for quantitative process control.
The Two-Pressure Reference Method
Both device types – GA315 and GA320 – operate exclusively according to the Two-Pressure Reference Method. This measurement principle is independent of cross-sensitivity compensation and humidity compensation and should not be confused with them:
- Measurement at normal pressure: The sample gas is passed through the measuring cell at approx. 1,000 mbar, and the IR absorption signal of the target component is detected.
- Reference at reduced pressure: The same sample gas is expanded to approx. 200 mbar. The concentration is determined from the difference between both signals – this ensures high zero point stability and robustness against intensity fluctuations and optical contamination.
Additional cross-sensitivity or humidity compensation through dedicated compensation channels (H₂O, CH₄, HC) is an independent mechanism and complements the measurement principle.
System Requirements for Hot-Wet Measurements
Regardless of the chosen technology, all components in the sample gas path must be heated:
- Heated sampling probe: At least 10–15°C above process temperature. GA315 for hot exhaust gases >200°C: Ceramic filter probe, heated to approx. 200°C.
- Heated filter: Same temperature as probe.
- Heated sample gas line: Formaldehyde (FTIR/GA315): 120–180°C. Ethanol (GA320): 60–80°C. Regulation required.
- Heated analyzer input: GA315: 90°C; GA320: 70°C.
Comparison of Measurement Technologies
Formaldehyde in Asphalt Recycling Exhaust Gases
| Technology | Detection Limit | Selectivity | Investment | Maintenance |
| FTIR (Hot-Wet, 180°C) | 0.1–0.3 mg/m³ | Very High | High | Medium |
| CRDS (Laser) | Sub-mg range | Very High | Very High | Low |
| NDIR Hot-Wet GA315 (HC Compensation) | ~2 mg/m³ (limited) | Good with comp. | Medium | Medium |
| Conventional NDIR with Gas Cooler | ~2 mg/m³ + washout losses >50% | Low | Low | Low |
| EC Sensor | Not suitable (Hot-Wet) | – | – | – |
Ethanol in Yeast Fermentation Processes
| Technology | Detection Limit | Selectivity | Investment | Maintenance |
| NDIR Hot-Wet GA320 (Humidity Compensation) | Good (ppm range) | Good with H₂O comp. | Medium | Medium |
| Conventional NDIR with Gas Cooler | Good | Good, but 10–35% washout loss | Low | Low |
| FTIR | Very Good | Very High | High | Medium |
Economic Considerations
The investment costs for hot-wet systems are significantly higher than for conventional systems with gas coolers. However, this additional expenditure is absolutely necessary and pays for itself through:
- Correct measurement results without systematic errors
- Legally compliant emission monitoring (formaldehyde case)
- Reliable process control (ethanol case)
- Avoidance of faulty batches due to incorrect measured values
- Reduced maintenance by eliminating condensate management
For ethanol measurement in yeast production, incorrect measured values can lead to suboptimal process control. The cost of a single faulty batch typically far exceeds the additional cost of a hot-wet system.
Conclusion: Physics Trumps Wishful Thinking
Water-soluble components require hot-wet measurement. The attempt to work with conventional gas cooling leads to systematic measurement errors for both formaldehyde and ethanol – physics cannot be outsmarted.
However, the specific measurement technology must be chosen application-specifically:
- Formaldehyde in asphalt recycling plants: FTIR or CRDS (preferred); GA315 only for less critical limit values (15 mg/m³).
- Formaldehyde behind biogas engines: GA315 with H₂O and CH₄ compensation.
- Ethanol in yeast fermentation processes: GA320 with humidity compensation.
Fresenius Umwelttechnik supports you in selecting the appropriate measurement technology for your specific application.
Frequently Asked Questions (FAQ)
Why can’t water-soluble gases be measured with a gas cooler?
Water-soluble gases dissolve in the resulting condensate upon cooling. The percentage washout loss is independent of concentration under constant conditions. Typical losses: 30–60% (formaldehyde) or 10–35% (ethanol).
Is the washout loss lower at higher concentrations?
No. The percentage washout loss is independent of concentration under constant conditions – this follows directly from Henry’s Law. The absolute loss increases proportionally, while the relative loss remains the same.
Can an EC sensor be used for hot-wet measurements?
No. Electrochemical sensors are exclusively suitable for cool, dry ambient air. They are neither stable nor sufficiently selective for hot, humid exhaust gases.
What is the difference between GA315 and GA320?
The GA315 is a true hot-wet system for process gases above 200°C: sample gas input 90°C, measuring cells 85°C, only one front-side input, integrated control for heated sample gas line. The GA320 is a quasi-hot-wet system for humid gases in the lower temperature range: multiple inputs via heated solenoid valves (70°C), measuring cells up to 60°C. Both devices operate according to the Two-Pressure Reference Method.
What is the difference between the Two-Pressure Reference Method and humidity compensation?
The Two-Pressure Reference Method (measurement at normal pressure, reference at approx. 200 mbar) is the fundamental NDIR measurement principle of both devices and serves for zero point stability. Humidity or cross-sensitivity compensation is an independent, additional measurement channel (H₂O, CH₄, or HC) that mathematically corrects the measurement signal.
Which standard applies to formaldehyde measurements in asphalt recycling plants?
There is no specific VDI standard for this application. VDI 3862 Parts 2, 3, and 8 are standardized for internal combustion engine exhaust gases. For asphalt recycling plants, the measurement procedure must be agreed upon with the competent authority.
