An online viscometer from Hydramotion has solved an important problem in process control for a leading South African specialty chemical manufacturer.
The largest supplier of organic coagulants in the southern hemisphere, produces a range of specialty polymers and chemical management systems. At the company’s Durban plant, viscosity is a key control measure for determining the degree of polymerisation during production of polyamine flocculants used in water treatment plants to assist in liquid-solid separation processes.
The Challenge: Managing Polymerisation Without Real-Time Data
As a result of a prior strategic decision, the company started to manufacture rather than “buy in” polyamines. Richard Winter, Production Manager, recalls:
“Our first plant was manually operated, producing 5000 kg batches. Because the operators had no previous experience of such manufacture, the potential for error was high, especially in the start-up phase. Having had some experience of polyamine manufacture, I estimated we could lose up to 50% of batches in this phase."
"Considering this, the process we put together was as simple as possible to reduce the possibilities of error which would result in ‘off-spec’ or — worse still — unrecoverable batches. We also wanted to keep our capital outlay for equipment to a minimum at this stage in the company’s evolution.”
With this in mind, the company decided to use existing resources as much as possible for the first plant. Small tanks already available on site were used to hold the raw materials, which were gravity-fed into the reactor as required. In an attempt to reduce the scope for error, a sight glass was fitted to each tank to enable operators to measure the correct quantity of material for each batch. The company also adapted its procedures to reduce the effect of raw material concentration variation errors (mainly relating to initial amine concentration) in monomer production.
The Risks of Off-Spec Production
In the absence of continuous, real-time viscosity measurement, the only way to monitor reaction progress was to measure samples of the reactor contents off-line, using a rotational viscometer. When the polymerisation was judged to have proceeded far enough, it was terminated by the addition of water and hydrochloric acid.
“Although this procedure worked,” says Richard Winter, “it was technique-dependent and there were some significant uncertainties. Was the sample representative of bulk properties? We supposed it must be, and we never had any evidence to suggest that it was not, but we could never be absolutely sure. Then there was the issue of maintaining the sample at reactor temperature after sampling and during the viscosity determination. There were two potential problems here: a drop in temperature would give an increase in viscosity, while continuing polymerisation would cause the viscosity to increase further still.”
There were also practical difficulties. When taking samples manually, operators risked exposure to amine and epichlorohydrin vapours, even though they were using protective equipment. In the absence of a pump and recycle loop, significant quantities of sample had to be returned manually to the reactor. But the main problem was that manual sampling, testing and cleaning the equipment could take up to 15 minutes, “which was a long time to go without information” when the rate of polymerisation was fast or accelerating.
The problem was particularly acute when the end-point of polymerisation was being approached and constant sampling was required. Lack of information at this stage could result in the “kill point” being missed, which meant the upper limit of the required viscosity range would be exceeded.
“Sample taking had a definite technique,” recalls Richard Winter. “There was always the risk that the sample would ‘skin over’ if it cooled too much, particularly during the latter stages of polymerisation. If this happened a new sample would have to be taken, which meant we might not have any results for up to 30 minutes.”
Although the company attempted to reduce the error by graphing the results and extrapolating the time to the end-point, they still faced the possibility that up to 30% of the batches produced could be outside specification. The potential for this was borne out by experience during the first 10 batches.
The company feared that, even once operating experience had been gained, up to 10% of throughput could be “off-spec”, especially when the higher viscosity products were being manufactured. For minor deviations from specification the batch could be recycled, but if the batch “gelled” it could not be recovered. Waste handling then compounded the cost of the lost batch.
“We did not want leave ourselves open to such costs,” Richard Winter says. “We realised that an on-line viscometer was the answer, but what type should we select? Our requirement was for a robust instrument that we could simply bolt into the reactor without a temperature-controlled recycle loop. We wanted to avoid this, partly because it would have involved installing extra pipework, a pump and a control system to keep the polymer at the right temperature, but mainly because of the viscosity monitoring problems we would have had if this equipment malfunctioned."
"We were looking for a viscometer which could be fitted directly into our reactor and give us consistent readings that we could interpret to allow us to manufacture to specification with minimal fluctuation. The Hydramotion XL7 on-line viscometer promised to be such an instrument. We had no experience with such devices, but the downsides of running the process with the off-line sampling system were so great that we decided to try the unit anyway.”
Immediate Results: Real-Time Data Transforms Production
Following the installation of the first viscometer in the reactor, the company quickly experienced positive results. Good correlations were obtained when the company compared the off-line viscometer readings with those obtained by the Hydramotion XL7 for every stage of the process for all the grades of polyamine they produced. The XL7 takes a measurement every 2 seconds and supplies the company’s SCADA system with both a fully configurable 4-20 mA analog viscosity signal and serial data (RS232 or RS422/485).
“We were able to identify the appropriate key control parameters in terms of ‘raw’ Hydramotion units to achieve on-specification product,” recollects Richard Winter.
“Operator acceptance of the equipment was good. Their workload was reduced dramatically during the polymerisation stage of the process, which allowed them to concentrate better on process control. It became feasible to consider ‘multi-skilling’ of the team to run further processes that were being considered in the future. ‘Off-spec’ production rapidly fell to considerably less than 1% of output and has remained at that level to date. A further benefit which the Hydramotion viscometer helped us achieve was an NSF listing for some of our polyamine grades.”
The manufacturer later expanded its operations, building a computer-controlled polyamine production plant and increasing batch size from 5,000 to 20,000 kg. With a Hydramotion XL7 viscometer installed in the new plant, “off-spec” production has remained negligible. The company has had occasion to celebrate its choice, as the XL7 has gone on to prove its long-term reliability. The viscometers have been in service for many years, yet maintenance requirements have been virtually nil.
“The instruments have performed flawlessly since they were installed,” reports Richard Winter. “We have experienced no significant drift requiring recalibration in that time to date. The units have been in continuous use for several years.”
The company has been equally impressed with Hydramotion’s after-sales support. “On the one occasion when a bracket detached itself from inside one of our reactors and damaged the viscometer probe,” Richard Winter recalls, “Hydramotion provided excellent, rapid and cost-effective service.”
The continuous real-time viscosity measurements from the XL7 are also providing fresh insight into the process, offering the potential for further productivity improvement. Although the plant is currently running at capacity, “our database is now adequate to exploit the information gained by using the viscometers to further optimise the process.”
The benefits of installing the Hydramotion XL7 were immediate and substantial. It is, says Richard Winter, “a truly ‘fit-and-forget’ instrument that enabled us to remove some of the art from polymerisation and replace it with science. The Hydramotion viscometer is the largest single improvement we have made in the manufacturing process and it continues to help us keep our products within specification and reduce waste.”