218 API STANDARD 521
Figure F.3—Approximate Flame Distortion Due to Lateral Wind on Jet Velocity from Flare Stack
The location of the flame center is quite significant when radiation levels are examined. Flame length varies with emission
velocity and heat release. Information on this subject is limited and is usually based on visual observations in connection
with emergency discharges to flares. Figure F.1 and Figure F.2 were developed from some plant-scale experimental
work on flame lengths covering relatively high release rates of various mixtures of hydrogen and hydrocarbons.
Several equations for calculating flame length and approximating flame tilt are presented in the literature [36, 47, 82, 139,151].
Each equation has its own special range of applicability and should be used with caution, particularly since the combined
impact of several factors (radiation, radiant heat fraction, flame length and center, and flame tilt) shall be considered.
The example in C.2.3 is another approach to calculating the probable radiation effects, using the more recent
method of Brzustowski and Sommer [36]. The principal difference between these methods is the location of the flame
center. The curves and graphs necessary to simplify the calculations are included in C.2.
There are other methods that can be utilized to calculate radiation from flares. More sophisticated models that
consider wind velocity, exit flare gas velocity, flame shape, and flame segmental analysis can be appropriate for
special cases, especially with large release systems.
Most flare manufacturers have developed proprietary radiation programs based on empirical values. The F-factor
(fraction of heat radiated) values used in these programs are specific to the equations used, and might not be
interchangeable with the F-factor values used in Equation (F.1). These programs have not been subject to review
and verification in the open literature. The user is cautioned to assess the applicability of these methods to their
particular situation.