From Uncertainty to Confidence: Evaluating Flare Performance Using CONVERGE

If you’ve ever driven past an oil and gas facility, you’ve probably seen those large flames burning in the distance. Those are flares, and they’re doing something important: industrial facilities use them to burn excess natural gas that cannot be used economically. This process prevents the natural gas from escaping directly into the atmosphere, helping to safeguard the environment. However, if the combustion efficiency of the flares drops, methane—one of the most potent greenhouse gases—escapes unburned.

The key measure of flare performance is called the destruction and removal efficiency, or DRE, which indicates the percentage of harmful compounds a flare successfully eliminates through combustion. In general, DRE is difficult to measure experimentally. Regulatory frameworks commonly assume a minimum DRE of approximately 98% provided the prescribed operating conditions are met, such as minimum heating value, visible flame, continuous pilot, and exit velocity limits. 

When Cimarron Energy tested their hybrid flare system, they recorded a DRE of roughly 99%, beating the industry standard. Good news, right? Well, it’s not quite that simple. The question that kept coming up was: How could they be sure that the number was accurate? The location where you take your measurements makes a huge difference. Industry best practices suggest sampling at a distance of about twice the flame length downstream, far enough that combustion is complete, but not so far that the reading gets diluted by surrounding air or contaminated by background emissions. Finding the right spot is critical, but testing multiple locations in the field is expensive and time-consuming, causing Cimarron to turn to computational fluid dynamics (CFD). They collaborated with Convergent Science to simulate their flare’s behavior, targeting three purposes. First, validate CONVERGE CFD software’s capability to predict flame shape under different and extreme operating conditions, e.g., an attached flame, a lifted flame, and flame blowout. Second, numerically study the DRE at different locations to determine the ideal measurement location. Third, validate that their flare achieved a DRE of 99%, as observed in their experimental study. In this article, we’ll explore how Cimarron used CONVERGE to shed more light on their hybrid flare performance.

System Design: How the Hybrid Flare Works

A flare system safely disposes of excess combustible gases released from industrial facilities by burning them, with carbon dioxide and water vapor as the resulting byproducts. The Cimarron hybrid flare, shown in Figure 1, receives gases from two sources: high-pressure (HP) natural gas produced during crude oil extraction, and low-pressure (LP) vapor released from storage tanks. The system features dual HP inlets: a 4-inch inlet, shown in green in Figure 1(b), which delivers gas through small tubes to create high-velocity jets for improved fuel-air mixing; and an 8-inch inlet (colored sky blue in Figure 1) that is connected to a ring-shaped manifold that distributes gas evenly through larger tubes. A 6-inch LP inlet is present alongside the HP gas from the 4-inch line.

Air enters from the bottom through a dedicated pipe and mixes with the gases. A cylindrical shroud at the top surrounds the flare, protecting the flame from wind while entraining additional air through its openings to the surroundings. The momentum of the fuel-air mixture creates suction that pulls in the extra air. The flame then heats this air, reducing its density and encouraging further air entrainment, leading to better mixing and cleaner combustion. 

Simulation Setup

Figure 2 shows the geometry and the domain that was simulated. The boundary conditions were defined to replicate experimental test conditions. A steady wind of 4.47 m/s is defined at the inflow of the domain. The simulation accounts for buoyancy to accurately capture the effects of hot gases rising under gravity. 

Although the hybrid gas flare is designed to operate with both the HP and LP inlets active, the simulation was intentionally set up to represent a specific test operating condition using only the 4-inch HP inlet, while the 8-inch HP line and 6-inch LP line remained inactive. The single-inlet configuration reflects the experimental test conditions, not any limitation in the simulation’s capability to handle multiple inlets. The fuel flowed at a constant flow rate of 0.154 kg/s. The air supply rates were adjusted to 30%, 50%, 100%, and 150% of the stoichiometric air for the supplied fuel amount to test different combustion scenarios. The torch was simulated as a continuous high-temperature inflow with a flow rate of 0.0005 kg/s at a constant temperature of 2000 K. The species composition was defined assuming complete combustion of a stoichiometric methane-air mixture.  

CONVERGE’s SAGE detailed chemistry solver was employed to model combustion. SAGE uses local conditions to calculate reaction rates based on the principles of chemical kinetics. The SAGE solver accurately models combustion with high fidelity and reliably predicts critical outputs such as temperature fields, flame shapes, and emissions including NOx, unburned hydrocarbons, and carbon monoxide. Additionally, CONVERGE features different turbulence models to capture the effects of turbulent flows. Precise methods for modeling flow dynamics and turbulence are necessary to get a clear picture of what’s happening during the combustion process. CONVERGE also features fully autonomous meshing, which saves users meshing time by automatically generating the mesh at runtime. Furthermore, CONVERGE’s Adaptive Mesh Refinement (AMR) refines the mesh when and where necessary, reducing computational time while maintaining high mesh resolution in critical areas. In this simulation, AMR determines where additional cells are needed based on the spatial secondary derivatives of temperature, velocity, and CH4 mass fractions. With the defined mesh settings, the largest cells in the domain are 400 mm, and the smallest cells—in and around the flame front—are 12.5 mm.

What Did the Simulations Say?

With 30% stoichiometric air, the flame adopts a conical shape and is tilted in the direction of the wind. As shown in Figure 3, the flame stayed attached to the torch tip and remained stable. The wind didn’t disrupt combustion; instead, it created natural mixing that pulled in air around the flame. From the simulation, a DRE of 99.96% was obtained, matching Cimarron’s experimental findings. The DRE was measured at a distance of 14 m downstream from the flare tip, where steady-state conditions are observed. As shown in Figure 4, the DRE does not change significantly beyond this distance, indicating minimal oxidation occurs after this point.   

Figure 5 shows the impact on flame stability and DRE for different stoichiometric air supply conditions: 50%, 100%, and 150%. At 50% air supply (Figure 5(a) and 5(b)), the flame remained stable and attached to the shroud, but DRE dropped to 99.5%. At 100% air supply (Figure 5(c) and 5(d)), the velocity near the flare tip exceeded the flame speed, causing the flame to lift off and detach from the torch. While combustion continued due to fuel flow and wind effects, DRE dropped significantly to 47.5%. At 150% air supply (Figure 5(e) and 5(f)), the aeration becomes excessive, completely extinguishing the flame due to the high strain induced by the air velocity.

Benefit of CFD

To summarize, we simulated Cimarron’s hybrid flare, demonstrating that you can obtain high-fidelity results using CONVERGE, and predicted the flame shapes under extreme operating conditions. But, how did this benefit Cimarron? After all, they already performed experiments to determine their hybrid flare behavior and DRE under multiple conditions. So why invest in CFD simulations? The answer: to increase their confidence in their results and show that the 99% DRE they obtained is accurate, not a flaw in the experimental measurement. CFD simulations are a valuable method for companies to demonstrate the efficacy of their products to their clients, increasing trust and building brand credibility.

To dive deeper into the physics behind this simulation, check out the full paper: Prediction of Methane Destructive Efficiency of a Gas Flare Using CFD With Adaptive Mesh Refinement and Detailed Chemistry, presented at the 2025 AFRC Industrial Combustion Symposium.

Gas flares are just one application of CONVERGE—we continue to expand its capabilities to help engineers solve critical problems across industries. If you’re interested in learning how CONVERGE can help enhance your company’s portfolio, contact us below!