2016 International Conference on Energy Development and Environmental Protection (EDEP 2016)

ISBN: 978-1-60595-360-1

Analysis on Dispersion and Toxic Effect of Liquefied Ammonia Tank

Release Based on PHAST

Zhi-Chao ZHANGa, Hui LIUb, Shi-Mei SUNc,*

JiLin Jianzhu University, ChangChun, China
a b c
zzchao1981@163.com, lhsafety@163.com, sunshir@163.com
*
Corresponding author

Keywords: Liquefied Ammonia Release, Dispersion, Toxic Effect, PHAST.

Abstract. Liquefied ammonia storage under the condition of low temperature and high pressure
is widely used in industrial production. The leakage of liquefied ammonia tank can cause
personnel intoxicant, fire and explosion accidents. In the present research, we study dispersion
and toxic effect of liquefied ammonia tank in the case of an instantaneous release. Wind speed
and surface roughness length are influence factors of ammonia dispersion. Distance downwind
of ERPG levels and toxic fatality show a trend of increase as the wind speed increasing.
Influence distance downwind of ERPG-1 and toxic fatality increase while distance downwind
of ERPG-2, ERPG-3 and ERPG-3* tend to decrease with the increase of roughness length. The
analysis of results allows us to determine the dead zone, toxic zone and security zone.

Introduction
Liquefied Ammonia is an important industrial raw material and widely used as a refrigerant
in industrial production. In the case of leakage, it will cause environmental pollution and
personnel intoxicant accidents. An ignition source during dispersion may cause fire and
explosion accidents, leading to significant casualties and economic losses. Therefore, studies on
liquefied ammonia dispersion and toxic effect have multiple significance meanings like
dividing security zone and preventing fire and explosion accidents. A number of domestic and
foreign researchers have studied the dispersion of ammonia [1-3] and consequence simulation
[4-6]. Some companies have developed calculation softwares, such as SAVE II, SIGEM,
WHAZAN, EFFECTS, PHAST, RISKCURVES, SAVEMODE, STAR[7]. PHAST(Process
Hazard Analysis Software Tool) is developed by DNV (Norway Classification Society), it can
be used to estimate the magnitude of consequences, determine effects of modifications, prepare
contingency plans and promote safety awareness. The software includes the toxic module,
radiation module and explosion module, providing 10 kinds of release scenarios, such as
catastrophic rupture, leakage, pipeline rupture and so on. PHAST consequence is in the form of
graphics and tables. PHAST has a wide range of calculation and the calculation results are in
good agreement with the experimental data.
In this paper, dispersion and toxic effect of an instantaneous release of liquefied ammonia
tank will be discussed under the condition of different wind speed and surface roughness length
by PHAST.
The Model

Instantaneous Discharge Model

This model[8] is used to describe the instantaneous release of an entire vessel inventory due
to a catastrophic rupture. It comprises the stage of expansion from initial conditions to
atmospheric. The expansion energy and final velocity are given by:
E exp = h0 − hf − (P0 − Pa )ν 0 . (1)
u f = 2E exp . (2)
Where Eexp is expansion energy, h0 is exit specific enthalpy, hf is final specific enthalpy, P0 is
exit pressure, Pa is atmospheric pressure, ν 0 is specific volume, uf is final velocity.
UDM Model
The dispersion model used in PHAST is UDM (Unified Dispersion Model)[9]. It can models
jet, dense, buoyant and passive dispersion. An instantaneous release profile is a volume defined
by revolving the vertical cross section around vertical axis. The UDM instantaneous cloud can
be characterized by an “equivalent” cylindrical cloud with effective radius and effective depth.
This equivalent cloud is taken to move with the cloud speed and to have an equivalent top-hat
concentration equal to the centre-line concentration.
Toxic Effect Model
The toxic model[10] calculates the toxic dose, the probit number, the probability of death, the
integrated probability of death and the exposure duration of an observer to finite concentrations
of a dispersing cloud. Toxic load(L), probit number(Pr) and probability of death(Pdeath) and are
given as follows:
T
L =
∫ [C(x ,t ) ]dt .
0
N
(3)

Pr = A + B ln(L ) . (4)
1  Pr − 5 
Pdeath = 1 + erf [ ] . (5)
2  2 

Where c(x,t) is the concentration measured in ppm of poison at distance x, dt is duration

measured in min, pr is probit number, A, B, N is a constant related to the toxicity of the
substance.

Toxic Effect Criterion

Toxic effect distances of ERPG levels are calculated in PHAST. ERPG (Emergency
Response Planning Guideline) are values developed by AIHA. It provides three kinds of
concentration limit of different acute toxic chemicals under the condition of exposure time up to
1 hour[11]. ERPG-1is the maximum airborne concentration below which nearly all individuals
could be exposed for up to 1 hour without experiencing more than mild, transient adverse health
effects or without perceiving a clearly defined objectionable odor. ERPG-2 is the maximum
airborne concentration below which nearly all individuals could be exposed for up to 1 hour
without experiencing or developing irreversible or other serious health effects or symptoms that
could impair an individual’s ability to take protective action. ERPG-3 is the maximum airborne
concentration below which nearly all individuals could be exposed for up to 1 hour without
experiencing or developing life-threatening health effects. ERPG levels of ammonia are 25ppm,
150ppm and 750ppm, ERPG-3 has been changed to1500ppm in ERPG Values(2015)[12].

Results and Discussion

The simulation is performed in the case of instantaneous release of a liquefied ammonia tank,
the volume of 50m3, the storage temperature of 20˚C, the pressure of 1.93Mpa and the filling
rate of 50%. In this area, annual minimum, maximum and extreme wind speed are 3.5 m/s, 7m/s
and 12m/s respectively. Three kinds of surface roughness length are considered in this
simulation, 0.03m (open flat terrain; grass, few isolated objects), 0.5m (parkland, bushes;
numerous objects) and 3m (city centre with high and low rise building).
Results at Different Wind Speed
Fig. 1-Fig. 7 and Table 1 show the results of dispersion process and toxic effect assuming
surface roughness of 0.5m and Pasquill stability of D.

Figure 1. Footprint of 25ppm at120s.

Figure 2. Footprint of 25ppm at 300s.

 Figure 3. Centerline concentration and footprint of 25ppm at 420s.

Figure 4. Maximum concentration footprint of 25ppm. Figure 5. Maximum concentration footprint.

Figure 6. Outdoor dose footprint. Figure 7. Outdoor toxic lethality footprint.

Table 1. Distance downwind of toxic effect at different wind speed.

3.5/D 7/D 12/D
Wind speed[m/s]
Influence distance downwind[m]
ERPG-1[ppm] 4689.7 5706.1 6498.7
ERPG-2[ppm] 1523.6 1887.6 2413.2
ERPG-3[ppm] 1012.3 1180.9 1683.7
ERPG-3*[ppm] 851.3 1028.4 1336.7
Toxic fatality 450 480 510
Notes: ERPG-3* is 1500ppm in ERPG Values(2015)
Fig. 1-Fig. 3 show cloud footprint of 25ppm at different time and different wind speed. As
time increasing, the cloud advances and drifts downwind and the effect zone of equivalent
concentration tends to increase. The cloud footprint at the wind speed of 12m/s disappears at
420s, which because the centerline concentration decreases to 18.6ppm, lower than 25ppm. Fig.
4 and Fig. 5 show the shape of the contours of the maximum concentration reached during the
dispersion. Fig. 6 and Fig. 7 show dose footprint and toxic lethality footprint of the cloud.
 It can be seen from Table 1 that the distance downwind of ERPG levels increases with the
increase of wind speed. The maximum distance downwind of toxic fatality increases as the
wind speed increasing.
Results under Different Surface Roughness Length
Assuming wind speed is 3.5m/s and Pasquill stability is D. The calculation results are shown
in Fig.8-Fig.10 and Table 2.

Figure 8. Centerline concentration at 300s. Figure 9. Centerline concentration at 600s.

Figure 10. Centerline concentration and footprint of 25ppm at 900s.

Table 2. Distance downwind of toxic effect under different roughness length.

0.03 0.5 3
Roughness length [m]
Influence distance downwind[m]
ERPG-1[ppm] 4673.3 4689.7 5076.5
ERPG-2[ppm] 2076.5 1523.6 1520.9
ERPG-3[ppm] 1434.8 1012.3 751.5
ERPG-3*[ppm] 1214.3 851.3 661.6
Toxic fatality 310 450 720
Notes: ERPG-3* is 1500ppm in ERPG Values(2015)
As seen from Fig.8-Fig.10, the maximum centerline concentration under each roughness
length decreases as time increasing. At either dispersion time, the smaller roughness length has
the higher maximum centerline concentration. When time is 300s and roughness length is
0.03m, 0.5m and 3m respectively, the centerline concentration reaches 840.1ppm of distance
downwind 1008m, 185.2ppm of distance downwind 1137.8 m, and 127.3ppm of distance
downwind1717 m.
 Table 2 shows the toxic effect on distance downwind under different roughness length. As
roughness length increasing, distance downwind of ERPG-1level increases while distance
downwind of the other three concentration levels decreases. Influence distance downwind of
toxic fatality tends to increase along with the increase of roughness length.
Conclusions and Future Work
An instantaneous release of liquefied ammonia tank was modeled by PHAST in this paper.
Wind speed and surface roughness length affect dispersion and distance downwind of toxic
effect. As to wind speed of 3.5m/s, 7m/s and 12m/s, the higher wind speed is, the farther
distance downwind of ERPG levels and toxic fatality has. The maximum centerline
concentration of cloud decreases with the increase of surface roughness length at the same time.
Centerline concentration of cloud under different roughness length decreases as time increasing.
Distance downwind of ERPG-1and toxic fatality tends to increase while distance downwind of
the other three concentration levels increases as roughness length increasing.
Further research on other influence factors of dispersion is necessary to determine the distance
of concern concentration, divide security zone and prepare contingency plans.

Acknowledgement
This research was financially supported by the State Administration of Work Safety.

References
[1] Re´my Bouet, Ste´phane Duplantier, Olivier Salvi, Ammonia large scale atmospheric
dispersion experiments in industrial configurations, J. Journal of Loss Prevention in the Process
Industries.18 (2005) 512-519.
[2] Sun Dongliang, Jiang Juncheng, Zhang Mingguang, Study on liquid Ammonia Storage
Vessel Release and Diffusion Models Improvement, J. Industrial Safety and Environmental
Protection. 37(2011)27-40. (In Chinese).
[3] Nishant Pandya, Nadine Gabas, Eric Marsden, Sensitivity analysis of Phast’s atmospheric
dispersion model for three toxic materials (nitric oxide, ammonia, chlorine), J. Journal of Loss
Prevention in the Process Industries 25 (2012) 20-32.
[4] A.D. Galeev, E.V. Starovoytova, S.I. Ponikarov, Numerical simulation of the consequences
of liquefied ammonia instantaneous release using Fluent software, J. Process Safety and
Environmental Protection 91(2013) 191-201.
[5]Hua Min, Yin Xin, Pan Xuhai, MATLAB Analysis on Emergency Area Decision and
Poisoning Risk for Ammonia Release, J. Industrial Safety and Environmental
Protection.38(2012)5-6, 36.
[6]Luo Aimin, Shi Lichen, Duo Yinquan, Wei Lijun, Comparative studies on accident mode of
anhydrous ammonia leakage, J. Journal of Safety Science and Technology, 3(2007)21-24.
[7]Wu Zongzhi, Progress in the study of foreign risk assessment software, J. Science and
Technology of Labour Protection. 14(1994)24-29. (In Chinese).
[8] DISC Theory Document, DNV Technic, April 2011.
[9] UDM Theory Document, DNV Technic, October 2012.
[10]Model For Toxics Calculations, DNV Technic, October 2005.
[11] Wang Yinghua, Ni Chunhui, Liu Linfei, Chen Xiao-min, Division and Estimation of
Functional Areas in Emergency Response for Toxic Gas Tank Leakage, J. Occup and Health.
27(2011) 2065-2066. (In Chinese).
[12] Information on https://www.aiha.org/get-involved/AIHAGuidelineFoundation/
EmergencyResponsePlanningGuidelines/Pages/default.aspx ERPG Values 2015.