Abstract
Acid dew point is a critical parameter in boiler design and operation, directly impacting flue gas temperature, boiler efficiency, coal consumption, and waste heat utilization. This study evaluates common acid dew point estimation formulas based on sulfur content, $SO_3$ concentration, and water vapor content across ten typical coal varieties. Results indicate that while most formulas correlate with these parameters, the И.A. Bapahoba (Vapahova) and Japan Electric Power Industry Center models provide the highest accuracy relative to the A.G. Okkes benchmark. Conversely, the Former Soviet Union empirical formula tends to underestimate the dew point, potentially increasing corrosion risks.
Introduction: The Efficiency-Corrosion Dilemma
In coal-fired power plants, the combustion process generates sulfur oxides ($SO_2$ and $SO_3$) and water vapor. When flue gas temperatures drop below the acid dew point, these components condense into sulfuric acid ($H_2SO_4$), causing severe corrosion on boiler tail surfaces. Engineers face a delicate balance: maintaining high exhaust temperatures prevents corrosion but reduces thermal efficiency (a 15–20°C increase costs ~1% efficiency). Conversely, lowering temperatures improves efficiency but risks rapid equipment degradation.
Methodology and Data
To identify the most reliable predictive models, this study compared several international formulas:
- Empirical Models: Former Soviet Union, Japan Electric Power Industry Center.
- Semi-Theoretical Models: Muller curve, A.G. Okkes, Verhoff & Branchero, Halstead, and Haase & Borgmann.
These were tested against a dataset of ten Chinese coal varieties, including Lignite, Anthracite, Bituminous, and Lean coal. The testing parameters spanned 0.2% to 1.3% sulfur content, 3 to 12.9 ppm $SO_3$, and 6% to 15.1% water vapor.
Comparative Analysis of Results
The Role of Sulfur and $SO_3$
The analysis confirms that sulfur content and $SO_3$ concentration are primary drivers of the acid dew point. However, the relationship is non-linear and varies significantly across different mathematical models.
As shown in the figure above, different models yield widely varying estimates for the same sulfur content. The Former Soviet Union model consistently provides the lowest estimates, while the Halstead and Muller curves tend toward the higher end.
The sensitivity to $SO_3$ concentration is particularly acute at lower levels (below 8.5 ppm). Beyond this threshold, the dew point continues to rise, but the rate of increase diminishes, as illustrated in the comparative plot of the various formulas.
Model Performance Benchmarking
The A.G. Okkes formula is widely regarded as the gold standard due to its high precision (error < 1.5°C). When compared to this benchmark:
- Best Performers: The И.A. Bapahoba (Vapahova) and Japan Electric Power Industry Center formulas showed the minimal deviation.
- Underestimation Risk: The Former Soviet Union formula, though popular, consistently under-predicted the dew point. This is likely due to its inclusion of a fly ash absorption factor, which may not be universally applicable or safe for all coal types.
- High Error Models: The Halstead and Haase & Borgmann formulas exhibited significant errors, reaching up to 20% in some cases, making them less suitable for precise engineering applications.
Conclusion
Accurate estimation of the acid dew point is essential for optimizing boiler performance. While the Former Soviet Union formula is a common industry standard, this study suggests that the Vapahova and Japan Electric Power Industry Center models offer more reliable predictions across diverse coal types. Future research should further investigate the complex interactions between fly ash and acid condensation to refine these predictive tools.
Summary updated