Statement
uring construction operations, an outlet line on the bottom of a benzene storage tank is
D sheared off. The discharge continues for 30 minutes before it is stopped. The diked tank is located in the tank farm of a large refineqdchemical plant complex surrounded by urban and
Release Attributes
Material: . . . Benzene Fluid state: . . . Liquia Chemical reactions? . . . Nc
Method: . . . Evaporating pool
Release time type: . . . Steady state
“Cloud height: . . . Ground leve, Roughness type: . IndustriaWSuburbar:
Stabilify: . . . L Averaging time: . . . 10 min.
Hazard:. ... Toxic
suburban populated areas. The minimum distance from this particular tank through the complex to the company fence line is 195 meters. Tank size and information needed for discharge rates as well as dike dimensions are given.
Assume the potential release could occur in the morning of a warm day with full cloud cover.
Required are estimates of downwind maximum concen- trations of benzene as well as comparative plume areas shown by 1 and 50 ppm benzene concentration isopleths. The effect of wind speed is also of interest. Use a 10 minute averaging time for all concentration estimates.
Analysis
Sourcehtelease Parameters Table 8-1.
Because the normal boiling point of benzene, 353 Ky is below any possible ambient temperature, the evaporation from the liquid pool formed by the release will be by convective mass transfer into the wind. The mass transfer rates were therefore calcu- lated using Equations 3-1 19 through 3-122 to find the evaporative mass transfer coefficient for the constant area pool. Physical properties required for these calculations were obtained from Chapter 3 of Perry’s Sixth; see Table S8-I.
Table 8-2.
STORAGE TANK PARAMETERS Liquid Benzene Release
Tank diameter, rn 8.0
Initial liquid level, rn 5.0 Square diked area side, rn 11.6 Discharge pipe diameter, cm 7.62 Dike area less tank area, m2
(area for evaporation) 84.3
Tank height, rn 5.5
Height of dike, m 3.3
Temperature of contents, K 300
BENZENE PROPERTIES
Normal boiling point, K 353.25
Molecular weight 78.1 1
Liquid density, k@m3 871.8
Kinematic viscosity of Molecular diffusivity of
Vapor pressure, Pa 1.382E+04
(Following are for 300 K)
vapor in air, rn2/s (v) 1.53E-05 vapor in air, m2/s (6) 7.70E-06
The various tank and dike dimensions, along with the liquid level in the benzene tank, are given in Table 5’8- 2 .
The pressure corresponding to the liquid head 5.0 m was calculated to be 43.18 kPa (= 5.0m471.8 kg/m3
9.80665* m/s2/1000). Using the liquid orifice Equa- tion 3-74 for the flow from the 7.62 cm discharge pipe, the mass flow rate is 39.57 kg/s. This gives a volumet- ric flow of O. 04539 m3/s.
The liquid volume discharged in the 30 minutes is 81.7 m3 or 71,228 kg, which corresponds to a pool depth of
* This is the gravitational constant. Also, more significant figures are shown than can be justified so the calculations can be “tracked.”
I
Copyright American Petroleum Institute
--`,,-`-`,,`,,`,`,,`---
A P I P U B L * 4 b 2 ụ ợ b = 0732290 0 5 b 0 1 1 9 T I T
S8-2 Chapter 6
0.97 m. However, the time required to just cover the floor of the diked area with liquid would be very small, so steady state dispersion models were used.
For wind speeds (U) of 4.0 and 8.0 m / s (see later), the evaporation fluxes are I . 868.
kg/[m2*s], respectively, using the convective mass transfer equations. For the pool area of 84.3 m2, these values give corresponding total emission rates of O. 157 and 0.288 kg/s.
Other Parameters
The specified atmospheric conditions indicate neutral stability, Class D. With this the principal wind speed used was 4 m / s , and to demonstrate the effect of a higher value, 8 m / s was also used.
Two roughness parameter values were used. For the plant complex area, zr = 1.0 m was used with
HEGADAS to the 195 m distant fence line, then zr = O. 1 for the urban/ suburban area beyond. To show upper concentration bounds, a constant z, = O. 1 m was also used with this model. Because this parameter must be constant with SLAB and DEGADIS, zr = O. 1 m was used.
Simulation
and 3.42 I I MODELING PARAMETER RECAP Source Information
Benzene properties Table S8-1 Storage tank data Table S8-2 Liquid discharge rate, kg/s .
Liquid discharge rate, rn%
Liquid discharge time, s 1 800
Volume discharged, rn3 Release rates, kg/s
4 mis wind speed 8 m/s wind speed Source area, mz
Evaporation fluxes, kg/[m2*s]
4 m/s wind speed 8 m/s wind speed Atmospheric Conditions Stability Class
Wind speeds, mls Ambient temperature, K
39.6 0.04539 79 0.157 0.288 84.3 I .87-1 CY3 3.4240'
D 4,8 300
Averaging time, s 600
Rouahness Lenaths. rn HEGADAS (planthrroundings)
Breakpoint case 1.0/0.1
Uniform case 0.1/0.1
0.1
DEGADIS and SLAB, ali cases HEGADAS, DEGADIS, and SLAB steady state model
versions, with the above area source parameters, were used to estimate the benzene dispersion for the five cases noted in the legends in the figures. It was not necessary to correct the concentra- tions for downwind travel time averaging, because with a release duration of 30 minutes and a 4 m/s wind, the distance traveled by a cloud section is 7,200 m compared with the largest downwind distance of interest of about 1,500 m.
Figure 5'8-1 presents cloud centerline (maximum) concentrations vs downwind distances for the five cases, with corresponding plots for 1 ppm benzene isopleths in Figures S8-2 and S8-3. In the ủnt graph, the HEGADAS curve for the 1.0/0.1 zr case is significantly lower than for the constant zr = O. 1 m case; roughly half as large throughout the downwind distances.
Note that the 4 and 8 m / s curves for HEGADAS do not differ much although the release rate for 8 m / s is about 1.8 times that of the lower wind speed. The small difference between the curves is explained by the compensating inverse effect of wind speed on dispersed concentration.
--`,,-`-`,,`,,`,`,,`---
Scenario 8: Evaporating Pool of Liquid Benzene S8-3
Downwind Centerline Concentrations
Evaporating Benzene Pool
i 2
5 1 0 2
e
c z
O
c c e
O o1 O'
O C O N 0 :
Q1, n o . "-
O ZOO 400 500 BOO 1000 1200 1400
Downwind Distance, meters
Figure S8-1.
Concentra tion Is0 ple t h s
- nE00*5 z--O.lOm.4m/sGnd
- - - UEDDIS' z, = I .O/O.I m 4 - - ncw4z,=o.iorn.ak/innd
D E a ở 6 . u I O.lOm.4m/sGrd
... I.. ~ . ~ , ~ o . l o m . 4 n / , G n d
1 ppmüentene
i 2 5 Q c 100
O
a o t o
Y)
O U C
-
c - z Y)
t " "250 "Downwind Distonce, meters "500 " ' ~. . 750 I ". 1000 " " ' " 1250 " 1500 i
~~
Figure S8-2.
Concentration Isopleths
O 20 40 60 80 100 120 140 160 1ò0 200 -201 ' ' ' ' . ' ' ' ' ' ' ' ' ' ' ' ' ' ' '
Downwind Distance, meters
Figure SS-3. Comers and straight segments are caused by con- necting the models' output points with straight lines without smoothing.
Copyright American Petroleum Institute
--`,,-`-`,,`,,`,`,,`---
A P I P U B L * 4 6 2 8 96 0732290 0560222 678