24-MMP-B4 Mine Ventilation and Occupational Hygiene · May 2014
Question 4 of 6: Downwind LFL Distance for an LPG Leak
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Notes on this paper
National Exams (BC), 09-MMP-B4 Occupational Health, Safety and Loss Management
(Mine Ventilation and Occupational Hygiene), May 2014, 3 hours, open book with calculator
permitted. Answer any five of the six questions; every question (1-6)
is answered in full as a complete study resource.
Reference texts: Crowl & Louvar, Chemical Process Safety:
Fundamentals with Applications, 4th ed.; ACGIH, TLVs and BEIs and
Industrial Ventilation: A Manual of Recommended Practice; OSHA 29 CFR 1904
Recordkeeping; WorkSafeBC/BC Health, Safety and Reclamation Code for Mines.
Question 4: Downwind LFL Distance for an LPG Leak (20 marks)
Given. Liquid LPG leak rate $Q_L=0.23\ \text{m}^3/\text{s}$;
liquid density $\rho_L=450\ \text{kg/m}^3$; vapour density $\rho_v=1.8\ \text{kg/m}^3$;
LFL $=3\%$ by volume; wind speed $u=11\ \text{m/s}$; ground-level release, rural,
Pasquill–Gifford stability Class F (stable); leak duration 5 minutes (undetected).
Find. The downwind distance beyond which the ground-level
centreline vapour concentration falls below the LFL.
Approach. Convert the liquid leak rate to a total mass emission
rate (worst case: the released liquid fully flashes/evaporates to vapour), convert the
LFL to an equivalent mass concentration using the given vapour density, then solve the
steady-state ground-level Gaussian plume equation for the downwind distance $x$ where
the centreline concentration equals that value, using the Pasquill–Gifford Class F
rural dispersion-coefficient correlations.
Total mass emission rate. Treating the entire leaking liquid as
fully vapourising (a conservative worst case for a fire-hazard assessment):
$$\dot m=Q_L\,\rho_L=0.23\times450=\boxed{103.5\ \text{kg/s}}$$
LFL as a mass concentration. At the low dilutions relevant to a
flammability boundary, the mixture's molar volume is close to that of air, so the
partial mass concentration is well approximated by the volume fraction times the pure
vapour density:
$$C_{LFL}=y_{LFL}\,\rho_v=0.03\times1.8=\boxed{0.054\ \text{kg/m}^3}$$
Ground-level, centreline Gaussian plume equation. For a continuous
ground-level release with no reflection needed (source already at grade), rural Class F:
$$C(x,0,0)=\frac{\dot m}{\pi\,u\,\sigma_y(x)\,\sigma_z(x)}$$
with the Pasquill–Gifford Table 5-2 rural correlations for Class F:
$$\sigma_y(x)=\frac{0.04x}{\sqrt{1+0.0001x}},\qquad \sigma_z(x)=\frac{0.016x}{\sqrt{1+0.0003x}}\quad(x,\sigma\ \text{in m})$$
Solve for $x$ by bisection. Setting $C(x,0,0)=C_{LFL}=0.054\ \text{kg/m}^3$
and solving numerically for $x$ (the equation is transcendental in $x$ through
$\sigma_y,\sigma_z$):
$$x_{LFL}\approx\boxed{303\ \text{m}}$$
at which point $\sigma_y\approx11.9\ \text{m}$, $\sigma_z\approx4.6\ \text{m}$.
Check against the leak duration. In 5 minutes at 11 m/s the leading
edge of the plume has travelled
$$x_{travel}=u\,t=11\times(5\times60)=3300\ \text{m}$$
well beyond $x_{LFL}=303\ \text{m}$, so the plume has fully established a quasi-steady
concentration profile out to the LFL boundary well within the 5-minute undetected
window — the leak duration does not limit the hazard distance found above.
Fig. 4 — Ground-level Gaussian plume from the LPG leak;
the flammable envelope extends to roughly 303 m downwind before the centreline
concentration falls below the LFL.
Quantity
Value
Total vapour mass emission rate
103.5 kg/s
LFL mass concentration
0.054 kg/m³
Downwind LFL (unsafe fire-hazard) distance
≈ 303 m
A leak of this size, dispersing under stable (Class F) conditions where the plume
stays narrow and concentrated rather than spreading and diluting quickly, creates a
flammable zone extending roughly 300 m downwind — large enough to reach well
beyond the tanker's own footprint and into surrounding roadway/facility areas, which is
exactly why stable, low-mixing conditions are the worst case for a ground-level
flammable-vapour release even though they are often associated with calmer weather.
Check: assumes the entire liquid leak rate instantaneously
flashes to vapour (a conservative worst case for LFL reach) and that the standard
low-dilution approximation $C\approx y\rho_v$ holds at the LFL boundary; the problem's
own instruction to ignore dense-gas effects is taken at face value even though a real
LPG release (heavier than air as a cold, unignited vapour) would typically require a
dense-gas dispersion model for a fully rigorous hazard distance.