This website contains several interactive web pages (AlWeb). On these pages you execute evaluations and optimisation calculations concerning the
Hall-Héroult electrolysis cells to produce aluminum.
Introduction to AlWeb
Interactive Web Pages
Anode Table Layout
AlWeb User's Guides
AlPrg consists of several computer programs for PCs with the Windows operating system. Similar to the interactive web pages (AlWeb) the modules of AlPrg executes calculations
concerning the aluminum production by Hall-Héroult electrolysis. You may download, install and run this software on your own computer.
Introduction to AlPrg
AlPrg Modules User's Guides
Parallel to the interactive web pages (AlWeb) and the Windows computer programs for your PC (AlPrg) you may use corresponding Android applications for your tablet PC.
As AlWeb and AlPrg these apps executes the same calculations and evaluations concerning the aluminum production by Hall-Héroult electrolysis.
AlApps User's Guides
On several web pages you find the theoretical background i.e. the ideas, methods and relations AlWeb, AlPrg and AlApp are using to execute their task. The website contains also the
corresponing literature references.
Table of Contents
1. Principles Hall-Héroult
2. Mass Balance
3. Properties of the Electrolyte
4. Cell Voltage
5. Energy Balance
6. Current Efficiency
7. Raw Materials
8. Cell Operation
9. Fluoride Evolution Model
10. Environment, Emissions
You find on the web site a reference of the literature authors (name index) and a subject index.
Please look at the information about AlSem (Aluminum Smelting Seminar & Workshop).
Table of Contents
Fluoride Evolution Model
Fluoride evolution is often classified as gaseous and particulate. Gaseous fluorides are those that continue to be gases at ambient
temperature: HF, CF4, SiF4, and C2F6. Entrained and volatilized bath become particulate
at the lower temperatures above the cell. In addition, alumina and carbon dust will sorb HF and therefore contribute to particulate fluoride.
We follow here the Fluoride Evolution Model that W. Haupin and H. Kvande have developed and
published in several papers (Lit. 1984,
These papers define Fluoride Evolution as the fluorine (F) content (kgF/tonneAl or gF/kgAl) of the hydrogen fluoride (HF)
and of the particulate that leaves the the cell with the cell gas. Most of the evolution will be captured by the cell’s hooding and fume treatment system.
That which escapes capture will be called Fluoride Emission. Hence, fluoride emissions can be calculated from the
hooding efficiency (fhood: 0.95 – 0.98 for modern cells) according to Equ. 1.
Other sources of fluoride emissions include fluorides from bath adhering to spent anodes, fluorides from tapping ladles and fluorides from anode carbon
baking furnaces. This other sources of fluorides and the silicon fluoride (SiF4) are not included in the model. Only a small amount of SiF4 is produced.
It is hydrolyzed to silicon oxide (SiO2) and and hydrogen fluoride (HF) by the humidity (H2O) of the air. The aqmount produced depends on
the silica content of the anodes.
Schema Fluoride Evoltion.
The cell gas, CO2 and CO produced by electrolysis, is saturated with gaseous fluoride species (NaAlF4, its dimer Na2Al2F8 and NaF).
When the cell gas cools down these gaseous species condensate to solid fluoride particels. The humidity (H2O) of the alumina ore and the hydrogen of the anodes react with the electrolyte fluorides
to gaseous hydrogen fluoride (Primary HF Formation). Liquid droplets of the electrolyte are entrained to solidify when they are cooling down. And finally water of the moist air hydrolyzes partly the
solid fluoride particles to create hydrogen fluoride (Secondary HF Formation).
Fluoride evolution consists of particulate and gaseous fluorides. The particulate i.e. solid fluoride particles are formed by the condensations of
gaseous electrolyte species namely natrium aluminum fluoride (NaAlF4), its dimer [(NaAlF4)2 = Na2Al2F8]
and sodium fluoride (NaF) as well as solidified entrained liquid bath droplets. Gaseous fluoride is hydrogen fluoride (HF) that is generated when fluorides present in the
bath or in the vapour phase react with moisture i.e. water (H2O) of the alumina or the atmosphere
Another source of hydrogen fluoride is from the hydrogen content of the anode, either from adsorbed hydrogen or hydrocarbon
It was suggested that the hydrogen was initially oxidized by the anode carbon dioxide to form water, and this water subsequently hydrolysed the sodium
tetrafluoroaluminate to form gaseous hydrogen fluoride. However one can not distinguish between any of the proposed mechanisms, because the anode is
at a potential that enables both electrochemical oxidation of the hydrogen to water and the direct
formation of hydrogen fluoride.
HF generation from hydrogen within the anode and from water introduced to the electrolyte is associated with the Primary Generation
of hydrogen fluoride. The HF generation outside of the cell electrolyte i. e. the thermal hydrolysis of some of the particulate fluoride is termed as
Secondary Generation of hydrogen fluoride.
Most of the particulate fluoride evolved from cells results from vaporization of the electrolyte. The Haupin - Kvande Fluoride Evolution
model assumes that the gases produced by electrolysis (CO2 and CO) leave the bath carrying an equlibrium partial pressure of the bath species. As the the
cell gas temperature falls the vapour condenses to form particulate.
The Haupin - Kvande Model of 2002 uses a relation for the total vapor
pressure of the bath (VP) according to Lit. 2001:
and the model of 1993 applied the following equation:
The electrolyte vapor contains as major species NaAlF4, its dimer NaAlF4)2 = Na2Al2F8) and
sodium fluoride (NaF). The partial vapor pressure of sodium fluoride (PNaF) is given with:
For the equilibrium constant (Kp) of the dimerisation reaction
and solves the quadratic equation for the partial pressure of the monomer (PM):
and finally get the partial pressure of the dimer (PD):
To determine the specific masses of the gaseous fluoride species that leave the cell with the carbon dioxide and monoxide gase we determine how many moles of
CO2 and CO (ΣnkgAl) are produced by electrolysis. Using the expressions for the specific electrolytic productions of CO2 and CO
(SEPCO2 Equ. 2.20 and
SEPCO Equ. 2.21) we write:
By using the ideal gas equation we continue to determine the specific moles (nM) and specific mass
(mM) of the monomer NaAlF4 species
and finally the specific masses of the dimer Na2Al2F8 (mD), sodium
fluoride NaF (mNaF) and volatilized bath (FFP):
The cell gas entrains liquid bath as droplets. As the gas cools these droplets freeze and become particulate. The crust acts both as a filter to remove entrained liquid and provide a long exit path
for entrained droplets to settle out. The following equation takes care of this "catching" action with the factor fcatch. Haupin and Kvande
use fcatch = 0.9.
Haupin and Kvande developed an equation for the surface tension (σ) from data of Lit. 1983.
As already indicated in Equ. 2 the hydrolysis of aluminum fluoride (AlF3) produces gaseous fluoride as hydrogen fluoride (HF).
The moisture (H2O) reacts with AlF3 rather than cryolite (Na3AlF6) or any other bath component because the equilibrium constant for reaction of Equ. 16 is much higher.
The equilibrium constant for the reaction of Equ. 16 leads to the following expression for the partial pressure of hydrogen fluoride (PHF).
is expressed as linear function of temperature over the range of 1200 to 1300 K.
We calculate the partial pressure of water (PH2O) according to the Principal Equation
by considering the water content in the alumina ore and the hydrogen content of the anodes (Equ. 19A and 19B):
Simplifying and solving for the mole fraction of water gives the following expression for the partial pressure of water (PH2O) in the cell gas:
For the activity of alumina we use Equ. 4.3 of the chapter
Reversible Decomposition Voltage.
Activity data for AlF3 in the electrolyte were fitted by the equation:
To determine the gaseous fluoride generated by hydrolysis of bath (FGB in gF/kgAl) we could multiply like in Equ. 12
the partial pressure of hydrogen fluoride (PHF Equ. 17) with the atomic weigth of fluorine and ΣnkgAl. However
the calculated values showed to be larger than measured data. The reason is that hydrolysis does not proceed toward thermodynamic equilibrium. Equation 21 takes care of this kinetic behaviour.
Also a variable ffeed was added to take care of the alumina feeding technique. ffeed = 1 is used for point fed cells and ffeed = 0.5 for break and feed technique.
With the break and feed technique ore is dumped on the crust and later the crust is broken to add the ore. Part of the humidity is driven off as the ore lies on the crust. With point fed cells more of humidity enters the bath.
Another source of gaseous fluordie arises from hydrolysis of NaAlF4 vapor by moisture in the air brought in by the cell's exhaust draft:
Treating this reaction similar to the previus reaction (Equ. 17) gives:
The incoming air supplies moisture but also cools the cell gas, limiting hydrolysis. fhydr is an adjustable factor between 0 and 3
allow for variations in the kinetics resulting from changes in ore cover. fhydr = 1 represents average conditions.
The next relations determines the total particulate fluoride (FP), the total gaseous fluoride (FD) and finally the total fluoride (FT).
An adjustabe factor fcover accounts for poorly covered cells. For a good cover fcover = 1 for a poor cover fcover = 2 to 7 depending on how
large the hole or holes are on the crust.