Process Hengineering, You Say?

Article by Jimmy Hunter BSc CEng CSci

Animal feed from waste calcium hydroxide?

South Africa, 1975

Developing a monocalcium phosphate (MCP) process

THIS article highlights the fun and excitement of a successful novel process development with a team of highly-motivated technical people. This development also allowed us to solve a particularly difficult environmental problem at the same time, the ultimate aim of most process engineers!

South African chemicals group AECI produced phosphoric acid from phosphate rock and sulfuric acid at its Modderfontein factory near Johannesburg. At its factory at Umbogintwini (‘Twini), 400 miles away in Natal, it produced acetylene from calcium carbide, as part of a PVC plant, ie:

CaC2  +  2H2O  ---> C2H2  +  Ca(OH)2

The acetylene (C2H2) was then reacted with hydrogen chloride to form vinyl chloride monomer, from which polyvinyl chloride (PVC) is produced.

The byproduct, ie calcium hydroxide, or lime, was run into a dam, producing a dilute slurry with a high pH. This waste represented a serious environmental problem, since it would have to be neutralised before disposal. The plant had been running for years, so the dam was substantial.

Our boss, chemical engineer Bob Fogel, came up with the idea of producing monocalcium phosphate (CaHPO4), an animal feed, from phosphoric acid and waste calcium hydroxide, ie:

Ca(OH)2  +  H3PO4  ---> CaHPO4   +  2H2O

The challenge

The proposal was to site the full-scale plant at ‘Twini, since it seemed to be a lot easier to transport phosphoric acid there than to transport the dilute lime to Modderfontein. In addition, the ‘Twini site already produced a range of veterinary products, so an animal feed would fit in well with the product portfolio.

Our task, as process development group, was to produce a pilot plant for the process, to develop it and produce samples for potential customers. Fogel gave us 6 months to design, build and commission the pilot plant. We were happy to take on the challenge.

This sounds like a tall order, but we had some advantages. The group had originally been a fertiliser development group, so we had people with strong granulation experience, and equipment already set up for granulation, drying and screening. We decided to go for a reactor/granulator design. We had an existing pilot-scale reactor/granulator with the right materials of construction, and some suitable pumps, etc for the task.

We then wrestled with the classical process development siting problem, ie do we build the pilot plant at ‘Twini, and have our process development people travelling back and forth, or do we bring the lime to Modderfontein and run the pilot plant there?

Our main considerations were that transporting the pilot equipment 400 miles away would be expensive and the construction would be difficult to control. Moving people away from their normal place of work is costly, and also causes many problems for families and puts a strain on everyone. At Modderfontein, we would be able to control the construction and operation more closely. Our choice was therefore to transport the lime to M’fontein and develop the pilot plant on our site.

An important technical problem was the transport of the dilute lime from the dam in ‘Twini to Modderfontein. Since the lime concentration in the dam was only a few percent, very large volumes of dilute slurry would have to be transported. We looked at many alternatives, for example filtration and centrifuging. Finally, we hit on the idea of using a hydrocyclone. We found a hydrocyclone which would concentrate the slurry from a few percent up to 40%.

Progress

The next problem was how to transport enough thick 40% lime slurry 400 miles to Modderfontein. At this stage, Lady Luck smiled on us! We found that ‘Twini had recently “retired” an old 20-ton stainless steel tanker, which was sitting in its plant disposal yard! One of our technicians, Peter, was a qualified motor mechanic and had a heavy-duty licence. We flew him and one of our other technicians, Tony, to ‘Twini. While Peter was fixing the tanker, Tony was setting up the hydrocyclone, hoses and pump. A few days later, they drove to Modderfontein with a 15-ton load of 40% lime!

While waiting for Peter and Tony, the rest of the team had to work flat out to prepare the pilot plant for the new operation. Designing the reaction system was tricky, because it is a liquid/solid reaction, and hence strongly dependent on particle size. The product (MCP) is also insoluble. We used a wide-mouthed thermos flask to measure the kinetics and heat of reaction of the process and used the data to calculate the required rates into the existing pilot reactor/granulator. The effect of particle size on reaction rate is well documented (O Levenspiel, Chemical Reactor Engineering).  A simplified process flowsheet is shown in Figure 1.

The phosphoric acid (H3PO4) would be pumped continuously at a controlled rate into the reactor/granulator and the 40% lime slurry would be pumped continuously into the granulator at the same time. A recycle stream, after screening, would be metered into the granulator at the same time. The rate of the recycle stream would be controlled so as to control the moisture content and particle size of the product to the dryer.

The coarse product from the dryer would flow to a set of screens where the fines would be separated from the coarse particles. The largest particles would be milled continuously and fed back to the separator.

A stream of product of the correct size and composition would be removed continuously and bagged.

Figure 1: Simplified MCP plant flowsheet

We soon started to produce product of high quality.

The development was not all downhill. We found that the MCP product met all specifications except fluoride (F) content, caused by the high HF content of the phosphoric acid. We then had to develop a process for removal of HF from the acid before use, which took some time, but was successful.

A few weeks into the development, Fogel had a visit from a well-known continental engineering contractor who could offer MCP process technology. Fogel told him to come back in 6 months, in case we hadn’t completed the development. The contractor, with a thin smile, said that it would be “impossible to develop an MCP process in 6 months!”

Fortunately we were able to disappoint him. The full-scale plant is still operating, nearly 40 years later. The lime dam is now empty, so an alternative raw material source had to be found, which means that the environmental problem has also been solved. A fitting end to a great adventure!

Article by Jimmy Hunter BSc CEng CSci

Process engineering consultant, JHunter Process

Recent Editions

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.