Why the long-term stability of your infrared analyzer can be worth millions—and how manufacturers differ fundamentally.
Anyone operating a biogas plant, a wastewater treatment plant, or an emissions monitoring system knows the problem: the gas analyzer’s measured values are correct today—but will they still be correct in three months? In six? In a year? This gradual shift in measured values at a constant gas concentration is called drift, and it is the key quality criterion that separates good NDIR analyzers from poor ones. In this article, we explain why drift occurs, why it becomes so costly in practice—and how Fresenius Umwelttechnik achieves exceptional long-term stability with two different, complementary measurement principles.
What is drift—and why is it so insidious?
NDIR gas analyzers (non-dispersive infrared) measure gas concentrations by passing infrared radiation through a gas sample and evaluating the wavelength-specific absorption according to the Lambert–Beer law. The principle is physically robust—but the hardware ages. And that is exactly where the problem begins.
Drift refers to a systematic, slow shift in the measurement signal that has nothing to do with an actual change in gas concentration. It manifests as a gradual rise or fall of the baseline over days, weeks, and months. Unlike noise, which fluctuates randomly around a mean value, drift is a directional trend—and therefore particularly dangerous, because operators often only notice it once it has already caused significant measurement errors.
The main causes of drift in NDIR systems are:
Temperature and pressure fluctuations: Changes in ambient temperature affect the radiation characteristics of the IR source and the sensitivity of the detector. Pressure fluctuations change the gas density in the measuring cell and thus the effective absorption.
Aging of the IR source: Every infrared source loses radiant intensity over time. In systems without effective referencing, this loss of intensity directly affects the measurement signal.
Detector degradation: Pyroelectric detectors change their sensitivity over their operating life. Without a compensation mechanism, this creates another, difficult-to-predict source of drift.
Optical contamination: Dust particles, condensate films, and chemical deposits on mirrors, windows, and filter surfaces reduce IR transmission. In industrial environments—especially with biogas containing high H₂S concentrations—this effect is particularly pronounced.
The cost of drift: more than just incorrect numbers
In practice, drift is not an abstract measurement issue—it costs real money and jeopardizes compliance. Some scenarios:
Biogas feed-in: When feeding into the natural gas grid, the methane content determines the settlement price. A drift of only 3% on the CH₄ measured value in a plant with a feed-in capacity of 600 Nm³/h leads to systematic billing errors that can add up to five-figure amounts over months.
Process control: For process control in biogas plants, a drifted CH₄ or CO₂ measurement means an incorrect assessment of the gas composition. A methane content displayed too low leads to unnecessary interventions; a value displayed too high masks process problems—with direct impacts on the plant’s profitability.
Calibration and maintenance costs: Systems with a higher tendency to drift require more frequent calibrations. Test gas cylinders for CH₄, CO₂, and zero gas cost several hundred euros per calibration. With quarterly calibration of a multi-point system, five-figure sums accumulate over the device’s service life—costs that are eliminated or drastically reduced with a long-term stable measurement system.
In all cases: The cost of drift typically exceeds the investment in a long-term stable measurement system by a multiple.
Two approaches—one goal: eliminate drift at the source
Fresenius Umwelttechnik follows a clear philosophy in NDIR gas analysis: drift is not subsequently compensated by frequent calibration, but minimized from the outset through the device’s optical and electronic architecture. Depending on the application, we use two different, complementary measurement principles.
Approach 1: Dual-beam NDIR with reference detector
The dual-beam method (two-beam NDIR) uses an integrated optical beam splitter in the detector housing that evenly distributes the source’s IR light across all measurement channels—for example, the measurement channel for CH₄ and a reference channel. Because both channels use exactly the same light source via the same optical path, all changes—source aging, temperature effects, optical contamination—affect both channels equally. By forming the ratio of measurement and reference signals, these influences are automatically compensated.
Additional features of our dual-beam concept that further improve long-term stability:
- Power-controlled, thermally stabilized IR source: The radiant intensity is actively controlled and kept constant.
- Temperature-stabilized and encapsulated measuring chamber: The entire measuring cell is thermally stabilized and shielded from external environmental influences.
- Proprietary characteristic-curve evaluation: Our mathematical evaluation algorithms also take pronounced nonlinearities of the absorption curve into account—allowing us to precisely capture both very low and very high concentrations with a single calibration curve.
- Intelligent adaptive signal averaging: The system automatically detects whether a signal change is due to an actual concentration change or whether it
- is measurement noise or drift. Only genuine concentration changes are output as the measured value.
- Zero point adjustment without a test gas cylinder: In most applications, activated-carbon-filtered ambient air is sufficient for automatic zero point adjustment. Only in special cases is nitrogen required.
Calibration interval (dual-beam): 6 to 12 months—depending on the sensitivity of the measurement task and the environmental conditions.
Approach 2: The patented two-pressure reference method
The two-pressure reference method—also known as the vacuum reference method—is based on an elegant physical principle: the gas to be analyzed serves as its own reference.
The method runs in two successive measurement phases: In Phase I, the sample gas is at a defined normal pressure (p₁) in the measuring cuvette, and the infrared absorption is recorded. In Phase II, the pressure in the same cuvette is reduced to a significantly lower value (p₂, typically approx. 200 mbar absolute) by pumping out the gas. As the molecular concentration in the gas decreases with pressure, the absorption also changes. The gas concentration is calculated precisely from the ratio of the two absorption values.
The decisive advantage: Because both measurements take place sequentially in the same cuvette with the same emitter and the same detector, all slow changes—source aging, detector drift, contamination of the optical windows—affect both measurements in the same way. When forming the ratio, these influences cancel out completely. The result is outstanding long-term stability: long-term drift of less than 3% per year on the IR measurement channels.
No reference gas, no second optical path, no filter changes—the two-pressure method eliminates drift structurally, not through compensation.
Calibration interval (two-pressure reference method): Once per year—in practice, often the only external intervention required for precise, compliant measurement.
The comparison in figures
| Criterion | Fresenius Umwelttechnik | Standard single-beam/two-channel systems* |
|---|---|---|
| Automatic zero point correction | ZeroLock—no test gas, patented | Typically test-gas-based |
| Calibration interval (NDIR) | 1× per year | 3–6 months (manufacturer-dependent) |
| Device service life | > 30 years | 10–15 years (typical) |
| Ongoing test gas costs | Virtually zero (no test gas required for the zero point) | Ongoing (cylinders, pressure regulators, labor) |
*The figures in the right-hand column describe general market categories and do not refer to a single manufacturer. Specific specifications vary depending on model and configuration.
What does this mean in practice?
For operators of a biogas plant, a wastewater treatment plant, or a CEMS measuring point, long-term stability translates directly into operational advantages:
Less downtime. Each calibration requires an interruption of measurement operation. With systems that only need to be calibrated once a year, downtime is reduced to a minimum.
Lower operating costs. Test gas cylinders for CH₄, CO₂, and zero gas cost several hundred euros per calibration. With quarterly calibration of a multi-point system, five-figure sums accumulate over the device’s service life—costs that are eliminated with a test-gas-free zero point adjustment.
Reliable compliance. An analyzer that drifts by less than 3% over 12 months reliably meets compliance requirements. Uncertainty contributions remain small, which simplifies the entire validation process.
Proven longevity: Fresenius Umwelttechnik regularly services and recalibrates devices that have been in field operation since the 1990s. Spare parts are still available for these devices. A service life of more than 30 years is not a marketing promise, but proven practical experience.
Better process control. When the CH₄ and CO₂ measured values are correct, the assessment of gas quality is also correct—whether before grid feed-in, for CHP control, or for monitoring fermenter biology.
Drift is not a law of nature
The drift of an NDIR gas analyzer is not an unavoidable physical fate. It is the result of design decisions—the chosen optical architecture, the
referencing strategy, and the evaluation algorithms. Systems without effective drift compensation trade low acquisition costs for high ongoing calibration and maintenance costs, as well as the risk of significant measurement errors.
Fresenius Umwelttechnik takes a different approach: with the dual-beam method with reference detector and the patented two-pressure reference method, our NDIR analyzers achieve long-term stability that significantly exceeds the market standard. Less than 3% drift per year with the two-pressure method, without external test gas for the zero point, with a device service life of more than 30 years—this is not a marketing claim, but the result of more than 175 years of experience in analytical chemistry.
If you would like to know how our analyzers perform in your specific application, please contact us. Our measurement technology experts will be pleased to advise you—also on site.
Frequently Asked Questions (FAQ)
What is drift and why does it occur? Drift is a slow, directional shift in the measurement signal that has nothing to do with an actual change in concentration. Causes include, among others, aging of the IR source, changes in detector sensitivity, optical contamination of the measuring cuvette, and temperature and pressure fluctuations.
How do the dual-beam and two-pressure reference methods differ? The dual-beam method splits the IR light simultaneously into a measurement and a reference channel and compensates fluctuations by forming the ratio. It is particularly suitable for multi-component measurements. The two-pressure reference method measures the same gas sequentially at two different pressures in the same cuvette—this eliminates all sources of drift through mutual compensation, and no external reference gas is required. The resulting long-term drift is below 3% per year.
How often does a Fresenius NDIR analyzer need to be calibrated? Devices with the two-pressure reference method generally require only a single external calibration per year. For dual-beam systems, we recommend an interval of 6 to 12 months depending on the sensitivity of the measurement task. Zero point adjustment is automatic in both cases—in most applications without a test gas cylinder.
What happens if I calibrate my analyzer less frequently? With systems that have a higher tendency to drift, skipping a calibration can lead to systematic measurement errors—with the consequences described for billing, process control, and profitability. Our devices are designed to remain stable even over longer intervals. Nevertheless, we recommend annual calibration as quality assurance.
Is it true that Fresenius devices remain in operation for over 30 years? Yes—and this is not an advertising claim. Devices from the 1990s are regularly sent to Fresenius Umwelttechnik for annual maintenance and recalibration. Spare parts are still available for these devices. This is the best proof that high investment quality is the more convincing economic decision in the long term.
Which applications are your NDIR analyzers suitable for? Our NDIR devices are used, among other things, in biogas plants (CH₄, CO₂, H₂S monitoring), wastewater treatment plants, landfills, industrial processes, and environmental measurement technology. Please contact us—we will recommend the optimal measurement principle and the appropriate device configuration for your specific application.
