1. Introduction: Flow Measurement—An Unavoidable Hurdle in SCR Control

In Selective Catalytic Reduction (SCR) denitrification systems, flue gas flow rate is a critical input parameter for calculating urea/ammonia injection quantities. The essence of SCR control is a closed-loop system combining feedforward and feedback: the feedforward section calculates the base ammonia injection amount based on NOx concentration × flue gas flow rate × target removal efficiency, while the feedback section makes corrections based on the outlet NOx concentration. If the flow rate cannot be measured accurately, the feedforward calculation loses its foundation. The feedback loop not only has to bear a heavier correction burden but may also cause over-injection (ammonia slip) or under-injection (environmental non-compliance) due to response lag - the former corrodes downstream equipment and forms ammonium bisulfate, causing blockage in the air preheater, while the latter directly leads to environmental non-compliance.

However, flue gas flow measurement is precisely one of the most challenging aspects of SCR systems. Engine exhaust gas possesses harsh characteristics such as a wide temperature range, high humidity, particulate matter content, and flow velocity pulsation. Coupled with limited on-site space and severely insufficient straight pipe sections, conventional industrial flowmeters cannot be directly applied. Most systems adopt differential pressure measurement for cost considerations, but practical applications have exposed dual problems of “inaccurate measurement” and “inaccurate calculation.” Focusing on this pain point, this article systematically reviews the root causes of the problems, conducts a comparative analysis of mainstream differential pressure flow measurement methods, and explores feasible optimization paths.

2. Special Challenges in SCR Flue Gas Flow Measurement

2.1 Inherent Difficulties Caused by Engine Exhaust Characteristics

Whether from diesel engines or gas engines, exhaust gas exhibits the following characteristics, posing severe challenges to flow measurement:

Wide velocity range and severe pulsation: Engine operating conditions range from idle to full load, with exhaust flow rate variations exceeding 10:1. Obvious exhaust pulsation exists (especially in single-cylinder or few-cylinder engines), causing the differential pressure signal to exhibit low-frequency fluctuations that are difficult to read stably.

Large temperature variations: Exhaust gas temperatures range from approximately 200°C at low loads to over 500°C at high loads, causing the flue gas density to nearly double. Without temperature and pressure compensation, the flow error introduced solely by density changes can exceed 40%.

High moisture content: The volumetric fraction of H₂O produced by diesel combustion is approximately 8-12%, and can reach 15-20% in gas engines. High-humidity environments not only affect gas composition (thereby affecting density calculations) but also easily condense in pressure impulse lines, causing blockage.

Particulate matter content: Carbon soot particles (PM) from diesel engines will gradually adhere to the surface of the throttling element, changing the flow area and leading to flow coefficient drift; although gas engines have fewer particulates, they still contain small amounts of inorganic ash.

2.2 The Micro-Differential Pressure Dilemma Under Low-Load Conditions

This is the most core contradiction of differential pressure flowmeters in SCR applications. The output differential pressure of a differential pressure flowmeter is proportional to the square of the flow rate:

Q = C · A · √(2ΔP / ρ)

This means that when the flow rate drops to 30% of the full scale, the differential pressure is only 9% of the full scale; when the flow rate drops to 20%, the differential pressure is only 4%. Taking a diesel engine SCR system as an example:

Typical Calculation Example

The full-load exhaust flow rate of a certain diesel engine is approximately 1200 m³/h, and the orifice plate is designed with a full-scale differential pressure of 2500 Pa. At 30% load, the flow rate is approximately 360 m³/h, and the differential pressure is only 225 Pa; at 20% load, the differential pressure is only 100 Pa; at 10% idle, the differential pressure drops as low as 25 Pa.

The accuracy of conventional industrial differential pressure transmitters is typically ±0.1%~±0.25% of the full scale. If calibrated at a full scale of 2500 Pa, when measuring at the 25 Pa point, the error of the transmitter alone could reach ±2.5~6.25 Pa, with a relative error as high as ±10%~25%. Combined with factors such as pressure impulse lines and temperature drift, flow measurement in the low-load region is practically unusable.

2.3 Severely Insufficient Straight Pipe Sections

SCR systems are typically installed between the engine exhaust tail and the muffler, with extremely compact space. The exhaust pipeline often contains elbows, reducers, branches, and other resistance components, resulting in a highly non-uniform flow field distribution. All differential pressure flowmeters require a certain amount of upstream/downstream straight pipe sections to establish a fully developed flow velocity profile, but the site often cannot meet this requirement:

l Standard orifice plates require upstream ≥ 20D~40D (D is the pipe diameter) and downstream ≥ 5D~8D

l Venturi tubes require upstream ≥ 5D~20D

l Actual SCR exhaust pipelines often have less than 3D~5D from the elbow to the flowmeter installation point

Insufficient straight pipe sections lead to flow field distortion, causing the actual flow coefficient to deviate from the calibrated value and making measurement errors unpredictable. This is one of the important root causes of “inaccurate measurement” in SCR flow.

2.4 Lack of Temperature and Pressure Compensation—The Root Cause of “Inaccurate Calculation”

Current Status of the Problem

It is understood that a considerable number of SCR system manufacturers do not perform temperature and pressure compensation at all, directly calculating flow rate using operating condition differential pressure, or only applying rough temperature corrections. This is unacceptable in industrial processes—but it has long existed in the SCR field due to: cost sensitivity, limited space, and insufficient technical capabilities of some manufacturers.

Differential pressure flowmeters measure the operating condition volume flow rate, whereas SCR control requires mass flow rate or standard condition volume flow rate. Without temperature and pressure compensation, when the temperature rises from 200°C to 500°C, the flue gas density decreases by approximately 40%. Without correction, the flow rate reading will be approximately 20% lower (because Q ∝ 1/√ρ), directly leading to an underestimation of the ammonia injection amount.

2.5 Improper Installation Location—Flowmeter Installed Downstream of the Urea Nozzle

This is an erroneous practice that indeed exists but should not occur in the industry. In order to obtain a larger differential pressure signal under low loads, some manufacturers install the flowmeter downstream of the urea nozzle. The problems with this approach are very serious:

l Urea spray interferes with the flow field: After the urea solution (32.5% DEF) is injected, it undergoes a complex process of atomization → evaporation → pyrolysis → hydrolysis. The spray itself alters the flow field distribution within the pipeline, and the differential pressure signal no longer reflects the true flue gas flow rate.

l Interference from sudden temperature changes: Urea evaporation absorbs heat, causing a local temperature drop, which affects the density field distribution and further introduces errors.

l Risk of deposition and crystallization: If urea atomization is poor or mixing is insufficient, crystallization/deposition forms on the surface of the throttling element, gradually changing the flow characteristics and causing continuous drift of the flow coefficient.

l Logical contradiction: SCR feedforward control requires the flue gas flow rate before ammonia injection to calculate the injection amount. Placing the flowmeter after ammonia injection is equivalent to using the “altered flow rate” to back-calculate the “original flow rate,” which puts the cart before the horse in control logic.

3. Basic Principles and Systematic Limitations of Differential Pressure Flow Measurement

3.1 Basic Principles

Differential pressure flowmeters utilize the pressure difference generated when fluid flows through a throttling element to measure flow rate. According to Bernoulli’s equation and the continuity equation, for incompressible fluids (or low Mach number gas approximation):

Q_operating = C · ε · A · √(2ΔP / ρ)

Where:

l C - Flow coefficient (determined by the geometric shape of the throttling element and the Reynolds number)

l ε - Gas expansion coefficient (compressibility correction)

l A - Cross-sectional area of the throttling orifice

l ΔP - Measured differential pressure

l ρ - Flue gas density under operating conditions

Conversion to standard condition volume flow rate:

Q_standard = Q_operating × (P_operating / P_standard) × (T_standard / T_operating)

3.2 Three Levels of Systematic Limitations

The “inaccuracy” of differential pressure flowmeters in SCR applications is not due to a single cause, but rather the superposition of systematic problems at three levels:

The superposition of these three levels of errors can result in a total error exceeding ±30% under the most severe conditions, far exceeding the flow accuracy required for SCR control (typically ±3%~5%). This highlights the severity of the problem.