Chapter 2

Volume 2 Start Chapter 1 Chapter 2 Chapter 3 Chapter 4
 

2

Predicting the drift and spread
of chemical spills
 

2.1        Introduction

The drift and spread of a chemical spill in the aquatic environment should as early as possible after the start of the release be assessed or calculated so as to form a basis for a risk analysis. A simple, rough estimation is often better than nothing. The estimation should as far as possible be based on the spills physical properties as well as environmental conditions like temperature, wind, water current, etc.

There exist various computer models by which an operator after some education and training can elaborate a forecast of the spill’s future fate. However, it should be emphasized that the forecast’s reliability depends fully on 1) the model’s construction and validity, 2) how correct all input data are, and 3) how professionally the model is run.
 

Many computer models exhibit astounding limitations. It is usual that forecasting models for gas clouds are not able to consider the structure of the ground or water surface (e.g. flat country, forest, calm water, rough sea). Some models cannot even consider mountains as obstacles for the cloud drift.

 


Figure 2 - 1   The drift and spread of chemical spills
can be forecasted by computer models.


Some drift models are not able to account for the chemical’s physical properties (e.g. water solubility) which gives a misleading or erroneous picture of the their drift.

Below in Subchapter 2.6 a few examples are given on forecasting systems, just for the purpose of exemplification. The objective has neither been to try to find the “best” system, nor to evaluate or compare such systems.

 

2.2          Gas clouds

Forecasting the spread in air   (Figure 1-13, Method F1)

Applicable for Groups G, GD, E, ED, FE, FED, DE  (all Groups with G or E)

Forecasting the spread of gas clouds in air can be estimated very roughly for the Groups G and GD by means of  Figure 2 - 2. Such estimates should always be regarded with reservations and never be alternatives for monitoring.
 

 

Health Risk

Fire/Explosion Risk

 

Release

 

Ammonia,

vinyl chloride, chlorine

Methane (LNG), propane (LPG),

butane (LPG),

ethylene,

butylene-butadiene

Ammonia, vinyl chloride,

methane (LNG),

propane (LPG),

butane (LPG),

ethylene,

butylene-butadiene

 

tonnes

metres/

/nautical miles

downwind

metres/

/nautical miles

downwind

metres/

/nautical miles

downwind

0.1

1,000 / 0.62

200 / 0.12

200 / 0.12

1

2,000 / 1.24

400 / 0.25

400 / 0.25

10

5,000 / 3.11

1,000 / 0.62

1,000 / 0.62

100

10,000 / 6.21

2,000 / 1.24

2,000 / 1.24

1,000

20,000 / 12.43

4,000 / 2.49

4,000 / 2.49

Figure 2 - 2

N.B.

The above Figure 2 - 2 can also be applied to liquid chemicals (which are flammable or especially hazardous for health) in the Groups E, ED, FE, FED and DE. The spread of evaporated gas, from spills of these chemicals, can be calculated very roughly by multiplying the values in the table by VP/100, where VP is the liquid’s vapour pressure in kPa, which is less than 100 at ambient temperature.
 

It is often difficult to get time to calculate the spread of instantaneously formed gas clouds in accidents, even if handy computerized models are available.

Sometimes it is impossible to calculate and predict gas cloud distribution even with the aid of very sophisticated modelling tools. Certain atmospheric conditions and/or substance properties may result in peculiar gas behaviour that makes the forecasting difficult.

A good example of this is the following small chlorine accident (Figure 2 - 3) that occurred in very cold and calm weather. The stable, windless, atmospheric conditions caused very limited dispersion of the chlorine cloud, which moved a long way before dissipating.
 

 The graph shows the rough appearance of a very long (10-15 km) and narrow cloud of a release of 10 kgs of chlorine gas at -30oC and calm, stable wind conditions.


Figure 2 -3

 

2.3          Floating spills

It is complicated to forecast the behaviour of a chemical spill that floats on the water surface. The spill’s fate is influenced by the following processes:
 

a)

The drift on the surface

b)

The spread on the surface

c)

Evaporation

d)

Dissolution

e)

Chemical reactions and other conversion processes
 

Various laboratory models have been developed (e.g. Ref. 7), but very few have been validated against real spills under operational conditions at sea.

Simple forecasting models have been developed for spills of chemicals that float on the water surface. For the sake of simplicity the spills are supposed neither to evaporate nor dissolve. This principle can also be used for manual calculations and is briefly described below.

Forecasting the spread on water surface    (Figure 1-13, Method F2)

Applicable for Groups FE, FED, F and FD   (all Groups with  F)
 

Figure 2 - 4 shows how a floating chemical slick’s drift can be calculated by means of a vector diagram in the same way as oil spills. However, most chemical spills belonging to the above mentioned Groups, except for F, will disappear by evaporation and/or dissolution within roughly 10 hours.

 

 


Figure 2 -4

 

2.4          Dissolved spills in the water body

 

Forecasting the dispersion in water body
(Figure 1-13, Method P3)

The method described below is applicable to the Group D only.

If the current of the water body is slow and even the dispersion can be calculated very roughly according to Figure 2 - 5 and Figure 2 - 6. This method cannot be applied for stagnant (or almost stagnant water) or for chemicals where the density differ too much from that of the water. Nor can the method be used for very turbulent water.


Figure 2 -5

 

Concentration  1 g/m3

Concentration  1 mg/m3

Release   tonnes

a     metres

nautical miles

a     metres

nautical miles

1

500

0.3

5,000

3

10

1,000

0.5

10,000

5

100

2,000

1

20,000

11

1,000

4,000

2

40,000

22

Figure 2 - 6

 

2.5              Sinking spills
 

It is very difficult to calculate the fate of a spill that sinks to the bottom. The reason for this is the number of parameters that influence the process (cf. Figure 2 - 7).

 

The chemical’s density affects the velocity by which the chemical sinks to the bottom. Its surface tension and solubility (even if very low) influence its behaviour on the water surface as well as its dispersing and spread in the water body during its sinking towards the seabed. The water current together with the water depth and the chemical’s density have a decisive importance for how long distance the chemical will move in the current’s direction before it touches the bottom.

Picture source: P. Ashworth, UK

Figure 2 -7
A sinking chemical and its behaviours (Ref. 8)

 

The chemical’s duration on the bottom is among other factors dependant on its solubility. If the solubility is e.g. 1% or 0.001% it must obviously have a pronounced effect on its duration on the seabed. Also the existence of water currents close to the bottom influence the duration. The chemical may also penetrate into the bottom sediment. The degree of penetration depends on the sediment’s properties and structure.

 

2.6          Forecasting modelling systems

 

2.6.1             Introduction

 

There exist hundreds of  highly sophisticated forecasting modelling systems for prediction of the drift and spread of chemical spills. Many of them are highly theoretical and not so easy to use. It is a difficult task to find models that might be usable in an operational organisation. A few known systems have been selected as examples below in Section 2.6.2.

 

2.6.2             Examples of computerized modelling systems

 

 

Forecasting computer model

Name      ALOHA (Areal Locations of Hazardous Atmospheres)

                 http://response.restoration.noaa.gov/cameo/aloha.html
 

Application      Gases
 

Information

Emergency responders can use ALOHA to predict the behaviour of a chemical gas in the event of an accidental release.

ALOHA is a part of the decision support system CAMEO (Computer Aided Management of Emergency Operations) developed by US National Oceanic and Atmospheric Administration (NOAA) in cooperation with US Environmental Protection Agency (EPA). (Ref. 6)

 
Figure 2 - 8   Graphical ALOHA description of a gas cloud dispersion in air.

Properties
 Can predict rates of chemical release from broken gas pipes, leaking tanks, and evaporating puddles, and can model the dispersion of both neutrally-buoyant and heavier-than-air gases.
 

Advantages
 Free of charge.
Technical assistance is available.
 

Limitations
 Mainly aimed for gas releases over land under conditions where the wind speed is neither too low nor too high. Does not account for topographic effects. The earth is assumed to be flat and the mean wind speed and direction are assumed to be uniform at any given reference height.


 

Forecasting computer model
 

Name      MET (Modells für Effekte mit Toxischen Gasen)
 

Application      Gases
 

Information
In accidents when hazardous gases are released into the air it is not satisfactory just to calculate gas concentrations in order to make rapid assessment of health risks and safety distances. The reason is that inhalation of high concentrations during short time will give the same dose as lower concentrations inhaled over a longer period. Calculation of safety distances should therefore be based on both concentrations and emission or spill rates.

MET makes a dose-effect-coupling for effects of toxic gases and estimates risks of human injuries in the area in the wind direction of the accidental release.

The dose as integral of concentration versus time is a good criterion in a model, since it diminishes one important but uncertain source term, the emission time. But doses also are not significant enough, since there is a further toxicological step to the main aim, to estimate the effect of toxic substances on the people in the surrounding area.

MET consists of the following four main modules:

1.

The instantaneous release of toxic substances as a puff and the formation of a gas/air cloud mixture.
 
2.

The dispersion of the toxic gases and calculation of the concentrations as a function of the distance (half sphere box model).
 
3.

The transformation into doses.
 
4.

The dose/effect-coupling based on a modified pharmacological receptor theory to evaluate the health impact.

The input values that MET needs are: 1) escaped substance weight, 2) wind speed and 3) a threshold value for the substance. Other parameters are automatically provided by the system in order to calculate hazardous distances.

MET has modules for simulations of 1) the washout effect of the cloud by rain, 2) the influence of a simultaneous fire and 3) the dispersion characteristics of heavy gases.

The model is stable to the large variations of the toxic values, since it can integrate several different values. In addition the lower explosion limit is used to calculate the size of an explosive mixture of a substance and air. The effects on mixtures of substances e.g. from fires can also be calculated.


                                                                                          Picture copyright: MEMPLEX Keudel GmbH
 Figure 2 - 9   Graphical MET description of a gas cloud dispersion in air.


                                                                                      Picture copyright: MEMPLEX Keudel GmbH
 Figure 2 - 10   The data input menu of MET
 

Contact address:

ISi Technologie GmbH
Rorschacherstr.126
9450 Lüchingen
Switzerland
E-Mail: met@isitech.com
Website: www.memplex.com

 
Properties
 Can predict hazardous distances of chemical releases.
 

Advantages
 Technical assistance is available.
 

Limitations
Mainly aimed for gas releases over land under conditions where the wind speed is neither too low nor too high. Does not account for topographic effects. The earth is assumed to be flat and the mean wind speed and direction are assumed to be uniform at any given reference height.


 

Forecasting computer model
 

Name      CHEMMAP
 

Application      Floaters, dissolvers, and sinkers
 
Properties      Predicts the dispersion and fate of marine chemical spills.
 

Information
 CHEMMAP is developed by Applied Science Associates, Inc. (ASA),
Rhode Island, USA.

CHEMMAP predicts the likely trajectory and fate chemical spills in the marine environment. The system is particularly suited to contingency planning and emergency response for spills of chemical cargoes from ships but may be applied to any chemical discharge. The system contains GIS and a 3D spill model that predicts the movement of chemicals in the water. The model relies on environmental data such as wind and currents, physical data such as the proximity of shorelines, and chemical data that define the chemical's properties. CHEMMAP includes a biological effects model which evaluates the effects of chemical spills on fish, shellfish and wildlife.

CHEMMAP incorporates a number of model components including:

- simulation of the initial release and plume dynamics of a product
   lighter or denser than water
- slick spreading and transport of floating materials
- transport of dissolved and particulate materials in three dimensions
- evaporation and volatilization
- dissolution and adsorption
- sedimentation, resuspension and degradation

The model uses physical-chemical properties to predict the fate of a chemical spill. These include density, vapour pressure, water solubility, environmental degradation rates, adsorbed/dissolved partitioning coefficient (KOC), viscosity, and surface tension.

CHEMMAP has its own database of  900 chemicals with physical and chemical data properties. A software link is optionally available to a database of more than 40,000 pure substances and 75,000 common mixtures. The latter database also provide guidelines for how to determine the severity of the risk to health, how to handle a spill, how to store and transport chemicals, how to dispose of chemicals, what to do if a chemical catches fire and how to plan for an emergency response.

Figure 2 - 11 shows the modelling of an instantaneous release of benzene (10,000 metric tons) at the water surface.  The plume display is the Vertical Maximum Dissolved Concentration of Benzene in the water column (mg/m3) 40 hrs after the initial release.  The colour-coded legend is located to the right of the plume with a cross section showing the plume in 3-dimensions below the legend. Above the plume is a graph of the mass balance that displays the percent of chemical that has surfaced, evaporated, in the water column, in or on the sediment and what has gone ashore over time.

 


Figure 2 - 11                                 Picture source: Applied Science Associates, Inc.

Contact address:

Applied Science Associates, Inc.
70 Dean Knauss Drive
Narragansett, Rhode Island 02882-1143, USA
Tel: +1 401 789 6224
Fax: +1 401 789 1932
Email: asa@appsci.com
Web page: www.appsci.com

 

Advantages
 Integrated with a database of  900 chemicals. Includes a biological effects model which evaluates the effects of chemical spills on fish, shellfish and wildlife.
 

Limitations
CHEMMAP is so far (January 2002) not validated with any field tests but the developers plan to do so in the near future.

 

 

Forecasting computer model
 

Name      ChemSIS  (Chemical Spill Information System)
 

Application      Floaters, dissolvers, and sinkers
 
Properties      Predicts the dispersion and fate of marine chemical spills.

 
Information
 Provides information on the movement of spilt chemicals within the environment and their behaviour under the influence of wind, waves, current flows etc. Performs three-dimensional modelling of the chemical spill’s dispersion. Involves prediction of large scale evaporation resulting in a vapour cloud as well as insoluble chemicals that can sink to the seabed. Covers chemicals that  form surface slicks or disperse or dissolve within the water column.

ChemSIS has been jointly developed by BMT and AEA Technology plc. The new system is designed to form part of a tailored package of support for chemical spill and as such is supported by 24 hour cover (through the National Chemical Emergency Centre at AEA) and the availability of a wide range of chemical spill response services. ChemSIS, in common with all BMT's applications, has been developed under the Visual Marine Information Systems framework and therefore integrates directly with any existing systems such as oil spill or search and rescue (Ref. 50).


 Figure 2 - 12                                      Picture source: BMT Marine Information Systems Limited

A modelling of a spill of 30 m3 vinyl acetate where Figure 2 - 12 shows the spills trajectory after 3:20 hrs (spill size 230 x 1,818 m) and Figure 2 - 13 shows the degree of evaporation (26 m3) and dissolution (4 m3) after the same time and during intervals.


 Figure 2 - 12                                      Picture source: BMT Marine Information Systems Limited

Contact address: See Ref. 50.
 

Advantages
 The developers of  ChemSIS claim that it is the only chemical spill model available in the world that is validated under realistic field conditions.


 

Forecasting computer model
 

Name      3D Transport and Water Quality Model
 

Application      Dissolvers and sinkers
 

Properties
Calculates the drift and spread of chemical spills  in the aquatic environment. The model takes the following processes into account:
      • Spill volume and concentration
      • Water currents
      • Sinking, dispersion

 
Information
 

Figure 2 - 14 shows the modelled distribution four weeks after an experimental release of 100 tons of a low-toxic emulsifier nonyl phenol ethoxylate (water solubility appr. 10 g/l) in the sea off the Finnish coast in the Gulf of Finland.

The spill size is 20 km across and the predicted and verified concentrations range from 26 µg/l down to 2 µg/l at the lower edges of the spill. The release site is marked with X.
Figure 2 - 14

The ”3D Transport Model” has been developed in Finland by the National Board of Waters and the Environment and the Finnish Environment Institute (Ref. 9 and Ref . 10).

The ”3D Transport Model” is available at:

Finnish Environment Institute
P.O. Box 140
FIN-00251 Helsinki
 Finland

Operational contact point:

Maritime Rescue Coordination Centre (MRCC) Turku, Finland
Phone (24h): +358 204 1000
Fax: +358 2 250 0950

 

Advantages
 This model is primarily aimed for the evaluation of ecological changes in costal waters. In this context it is used for simple three-dimensional drift calculations of water-soluble chemical spills.
 

Limitations
When used by itself this model cannot predict the transport and spreading of substances due to the wind, waves, turbulence and sea currents. These parameters must be estimated or calculated by a three-dimensional hydrodynamic model.
     
 

End of Chapter 2