SOUTH Africa's Vaal River is the main source of water for the thirsty people and industrial plants in the main production area around Johannesburg.
The word “Vaal” is an Afrikaans word describing the colour khaki. The river is this colour because the water has a very high clay content, making it undrinkable. The Rand Water Board (RWB) knows how to solve this problem, using a dose of about 10 ppm of aluminium sulfate, followed by a settling process. The resulting water is drinkable, and suitable for most uses.
The aluminium sulfate (AS) market is huge, and has been controlled by AECI for many years. I previously wrote an article on debottlenecking of the AS plant, which was published in The Chemical Engineer1. The AECI process relied on the reaction between sulfuric acid and imported bauxite (aluminium oxide) as follows:
3H2SO4 + Al2O3 ––> Al2(SO4)3 + 3H2O
AECI has always jealously guarded the AS market. Our team's boss, Bob Fogel, noticed some significant movement in the mining sector in Namibia which had potential to disturb the AS market. A mining company had opened a large new flint clay (aluminium silicate) mine, and in partnership with other companies, had patented a new process for production of AS from flint clay. The new beasts in the jungle were out to win the AS market from under our noses.
Their process went as follows:
3H2SO4 + Al2(SiO2)3 ––> Al2(SO4)3 + 3H2O + 3SiO2
The patented process produced AS solution and solid silica.
Bob Fogel called me in and gave me my direction: I had to find a way to better our rivals' new process without using an unethical alternative process. We had to find a way to beat them at their own game without doing anything dishonest. This was a challenge we could not refuse or lose.
The patented process used a certain fixed ratio of acid to flint clay. We changed the process so that we always used an excess of acid, more than specified in the patent. This meant that we always had to use a two-step process, ie the excess acid always had to be neutralised. We used aluminium hydrate to neutralise the excess acid:
2Al(OH)3 + 3H2SO4 ––> Al2(SO4)3 + 6H2O
We realised that we had to move fast! We immediately ordered a suitable glass-lined stirred reactor to carry out the reaction experiments. A typical glass-lined reactor is shown below.
We had previous experience of reactions between solids and solutions. We knew that the particle size of the solids would be crucial, as pointed out by Levenspiel in Chemical Reaction Engineering2. We started reaction experiments as soon as the reactor arrived, which soon told us that we had a serious problem.
We started reaction experiments as soon as the reactor arrived, which soon told us that we had a serious problem.
The original process (reaction of bauxite with dilute sulfuric acid) produces a solution of AS and solids which settle easily and are easily separated from the pure solutions. The new process, using flint clay, produced a solid phase which did not settle, even after 24 hours.
We enlisted the help of an experienced filter company, whose representative had previous experience in dealing with hard-to-settle solids, and they proposed using special flocculants which would coagulate the silica solid and allow fast settling and filtration. We found a dosage of about 10 ppm which coagulated the solid silica and resulted in rapid settling. They were also able to rent us a rotary filter which could filter the solution continuously. Suddenly we had a process!
We connected the unit up to our pressure reactor and we were soon producing a clean, filterable solution of AS, with the added bonus of a dry silica cake which could easily be disposed of. We evaporated some of the AS solution and produced a clean solid which met RWB's specifications.
Our competitor, on the other hand, did not believe that RWB would insist on a product with a very low solids content. Unfortunately for them, they had fallen in love with their product and process.
We knew better! We had previously worked on developing a reaction/granulation process which would have been much cheaper than the existing process, but the RWB had been very stubborn about changes to its product specifications. We knew that we had to satisfy RWB in every respect. Staying close to our customer was the key to commercial success in this case.
A few weeks later, our team was invited to a demonstration at the competitor’s plant, along with the commercial people from our company. We noted that they were not using flint clay as the raw material, since the RWB had not approved the product.
This turned out to be the most bizarre plant visit we had ever attended. We realised that they were not able to offer the clean product because they did not have the technology which we had developed. However, we knew that we had won the battle for supply of AS from flint clay. We also realised that the visit was partly a face-saving exercise, so we had to be very polite and avoid any grandstanding. It was very difficult not to be triumphant, but to give them the respect they deserved.
Within a few months our company had taken over their plant at a very good price and conditions. We were still the biggest beasts in the AS jungle.
1. Hunter, J, "Debottlenecking Process Plant", The Chemical Engineer, March 2000.
2. Levenspiel, O, Chemical Reaction Engineering, vol 1.
To read more of Jimmy Hunter's Chemeng Chronicles, visit the series hub.
Catch up on the latest news, views and jobs from The Chemical Engineer. Below are the four latest issues. View a wider selection of the archive from within the Magazine section of this site.