Troubleshooting tales

A key skill of SN2 is troubleshooting plant that have failed or processes that have gone wrong. Our work in this area has contained costs by keeping our clients out of the courts or has resulted in a fix that has given a positive new direction.

Flowmeters Get the Jitters

Normally the rotating-gear positive-displacement flowmeters on the plant worked reliably. They were the type in which a magnetic mounted within the rotor tripped a micro relay on passing the stator. This gave a single pulse for every revolution which was read by a digital module and sent to the host computer for processing. Things went askew however when a variable-speed drive (VSD) was used to control the speed of the pump that delivered flow through the flowmeter. The flowmeter began to send an increased frequency of pulses, between 1.5 to 3 times the expected pulse rate. We looked at the pulses on the oscilloscope and found that the main pulse was accompanied by a much smaller pulse occurring just before or just after the main pulse.

We knew that the VSD was the culprit and we suspected that the fault had something to do with the pump being a positivedisplacement (progressive cavity) type. Unlike the wave form of a normal 3-phase supply, the VSD's output wave form is stepped and therefore not at all smooth. These steps impressed themselves into the torque of the electric motor and from there to the flow of water in the pipe downstream from the pump. The steps in torque became high frequency pulses moving at the speed of sound in water from the pump to the flowmeter. The result was that an extra little jiggle in the rotation of the rotor occurred when the micro relay was tripped and this was sufficient to add an extra pulse to the output. The spurious pulses were easy to damp out using a capacitor across the output. We never saw this phenomenon with centrifugal pumps and did not expect to do so.

Breathless diesel engines

Underground mining often uses trucks to move ore to the surface from deep below. When these 500-kwatt, turbocharged and intercooled trucks, worth about $1M per unit, misbehave there is anguish in the boardroom. Such was the case several years ago when failures were reported continually after only a few weeks in service. The diesel engines were known to be suffering from inlet valves' inability to close. The resulting symptoms were lack of engine power as the force of the burning fuel was leaked from an ill-fitting valve. Gas blew back into the inlet manifold when it should have been pushing the piston down. The owners were close to litigation when SN2 was asked to have a look.

The problem turned out to caused by the unlucky coincidence of several circumstances: the humidity in the mine, the characteristics of the intercooler, the sulfur content of the fuel and the type of explosive used in the mine. Light microscopy revealed a heavy layer of carbon on the valve seats and, unexpectedly, under that, a layer of copper metal where it should not have been. The explosive was AMPHO ammonium nitrate and diesel fuel. This always gives a level of ammonia gas after detonation. The high-sulfur fuel contributed sulfur oxides to the atmosphere in the mine which combined with the ammonia gas to make ammonium sulfate. The interesting thing here was that the ammonium sulfate was formed as an aerosol, with particle size in the order of tens of nanometers.

This salt was drawn in with the charge air and compressed by the turbo charger and then cooled in the intercooler. This is where the high humidity (essentially 100%) in the mine came into play. On cooling the air dropped out some of its water as liquid. Normally this is not a problem because the water enters the engine and is evaporated. But the water contained salts of ammonia and these could not be evaporated. Worse still the ammonia dissolved some of the copper in the intercooler and gave copper-ammonia chelates. At first salts solutions deposited in the inlet manifold where they were dried by the heat but over time there was an ineluctable creeping of salt solution to the inlet manifold. When the salts finally arrived it got coated onto the valve seats. The copper metal seen there was made by carbon reduction of the salt.

When management and the suppliers of the trucks got the full story, there was no more talk of litigation and the two parties worked on a solution.

Swimming pools go brown

An unusual problem emerged in the form of discolouration on a swimming pool gelcoat. Owners of pools threatened the pool builders with ruin if the problem was not addressed and fixed. The company had performed initial analysis on a scanning electron microscope that had failed to reveal the cause. Such is the multifactorial nature of certain problems that the answer turned out to involve microbiology, mineralogy and industrial relations or more prosaically: mould, rust and shonky contractors.

The binocular microscope revealed the gelcoat to have a crazed surface with refractile bodies lying in hollows. These bright crystals disappeared when treated with hydrochloric acid, the resulting washing being bright yellow strongly suggesting iron oxide.Another clue lay in the universal presence of mould in affected pools, especially evident on rubber articles in contact with the water.

Yet another clue was the use of yellow sand in the sand filters, clearly the source of the iron. A telephone call to the makers of the resin gave the last item in the list. Adding generous amounts of accelerator causes the gelcoat to crack after curing. This regretably widespread practice allowed contractors to finish a job in half the time and while the job might suffer microscopic cracks, these were not normally noticed. A piece of gelcoat was cultured in agar made with only sterilised poolwater as nutrient. Filaments of mould grew from the cracks suggesting that the mould was metabolising soluble iron.

From here we hypothesised that the mould sequestered iron from the water to coat itself in a chlorine-resistant iron-oxide layer that acted as a catalyst, in concert with sunlight, to convert harmful oxidising chlorine to useful oxygen. Using white sand in the filters ameliorated the problem. From then on stern warnings were given to contractors and a process of close supervision initiated. This probably increased the price of the job but ensured a reliable result and the continuing viability of the company.