For my third memoir, I go back to when I started working as an Operations Manager in my career in glass. I was working at Rockware Glass (now Ardagh Glass Packaging) in Knottingley, Pontefract, UK. At that time, I was assigned this role having come through the engineering route as Plant Engineer followed by Assistant Production Manager at the plant. In the new role, I would be taking [on] more core production activities, which included managing the departments directly to give improved quality and higher output.

Although I had already been inducted into production activities with my previous role of Assistant Production Manager, there were still many things for me to learn. However, a past involvement with forehearth control (see Parts 1 & 2 of these memoirs for details) gave me a head start on this aspect of the glass container-making process.

Management by walking around

One of my daily routines as an Operations Manager was to walk the full plant from the batch house to palletising. It might not have been in the job description as such, but I always believed in MBWA (Management By Walking Around) after being taught this by one lecturer at university. This routine walk commenced straight after we had held the shift handover meeting from the night shift. That meeting at least armed [me] with some specific matter to investigate, on most occasions.

When I reached the furnace control room on my daily walk, I was always amazed at how the level of understanding regarding glass melting pretty much stopped at the throat of the furnace. It was clear the furnace management team understood what to do to control the temperature and quality of the glass at the ‘riser’ (the exit of the throat into the refiner/distributor), but then, after that, my experience in forehearth control made me realise something was lacking in their knowledge.

It might be hard to believe, but even with the many thermocouple measurement points we have and the set-point process controllers we use on forehearths, the forehearths cannot easily correct for glass disturbances entering the forehearth. I therefore had to make it a point of education with the furnace department that this was the case. That was not going to be easy to communicate, I thought.

Educating the furnace department

The problem with the control of gob temperature is that we do not have a good measurement of the actual temperature homogeneity of the gob. We can observe the gob shape as the glass exits the orifice, and measure for any changes in gob length (using a gob speed monitor during gob delivery), but we lack certain other details.

In our case [at Rockware Glass], as is still the case in many plants, the exact measurement of the glass temperature ended at the exit of the forehearth in the transition channel block, before the glass enters the spout. (Currently, the best measurement is the 9-point thermocouple matrix, which many of the well-equipped plants already have.)

Therefore, even though the temperature readings that the furnace department had on their monitors for the forehearths may have looked stable, the gob homogeneity and perhaps average temperature could change. (We did not have any online optical measurement even of the surface of the gob in those days, although I know some plants have those now). It was therefore often the case that after a disturbance to the glass conditioning, it was only noted by a change in gob shape or the unevenness of the bases of the bottles, where the glass distribution had altered. It was not readily seen in the furnace department on their control monitors.

In the opinion of the furnace department, they had not caused the disturbance that we had seen on the production floor, even though there might have been a furnace upset such as a problem with glass level control, which was the most common disturbance we had back then.

My mission, therefore, was to at least try to educate the furnace team about what they could not see. Also, to teach them that this was causing production problems and the answer was to keep the furnace operating conditions more stable to assist with consistency in production performance. The question that remained was… how could I do this?

I had to appreciate that they had not had the level of exposure to daily monitoring of forehearths that I had benefitted from. Nor had they worked with a forehearth expert to learn from their years of experience. (Again, see Memoirs 1 & 2 for the details of that.)

Practical teaching aid

One day, I realised that there is one control that we had on the furnace control room monitors that not all plants have access to. That measurement is the measurement of the feeder tube height. I believe this should be a standard measurement in all glass plants making glass containers, but my experience is that this measurement is rarely available. Certain companies believe in it, but it is not a universally accepted essential ‘glass’ measurement point.

In the end, it was quite easy to convince the furnace team that the tube height measurement was useful to check if any disturbance to the glass had been caused by monitoring the tube height. The reason for this was simple: they understood it because of their familiarity with the properties of glass in terms of glass temperature and viscosity. They knew that if the glass temperature or glass level changes, then to keep the same gob weight, the tube height would need to be altered.

Other things affect the glass conditioning, but as far as the furnace department is concerned, the glass temperature and glass level control are the main factors that the furnace department can influence, so I kept the discussion around those. I explained that if they track the tube height of the forehearth it will tell them how stable the glass temperature and glass levels is that they are delivering to the forehearths. Of course, a change in glass composition would also give such an impact, but thankfully such disturbances were rare, as confirmed by daily glass density measurements.

Additionally, I had to explain that the feedback control loops on the forehearths were exactly that, feedBACK. There was no effective FeedFORWARD control and so any disturbance had to be seen (and delivered to the feeder) before it could be corrected for by any adjustment of the controls. That being the case, even slight disturbances could affect the stability of the gob temperature, shape, and weight (although the production operators always have to take action to compensate for that).

It is because the production department has to make compensation to maintain the weight that we can see the effect of any furnace disturbance on the forehearth as evidenced by the tube height variation. It might be a crude measurement but at least it is a measurement. It is also something that many plants don’t have (as I previously stated but repeat here for emphasis) and, in that sense, those plants are running ‘blind’ for this signal of disturbance.

There are automatic control systems to correct the gob weight and keep it stable. Most commonly this is used in NNPB but also now in BB and WM production where the benefits of more stable weight control have been recognised as well. However, even where there is no automatic weight control in BB and WM production, the operator will be the one taking the corrective action to the gob weight. This means that in all cases a change in tube height will still be seen in response to a change in glass conditioning that has been caused at the feeder.

Manual and supervisory control

It was not too long before the furnace department was quickly calling up the tube height control signal on their screens to show me the amount of disturbance caused, or not caused, at the feeders, following an upset in the furnace. They also accepted that forehearths were not fully automatic after all. In fact, I believe the furnace department should think of forehearths as though in manual control and keep the glass going into the forehearths as stable as possible.

Since that time, forehearths have increased in the complexity of control quite significantly. Nevertheless, the basic idea is the same and keeping stability going into the forehearths will give more consistency in glass conditioning coming out at the gob feeder.

As the glass industry continues to strive to become more efficient, forehearth control is still one of those areas that would benefit from further developments of supervisory control. However, the need for that is somewhat lessened when the input to the forehearth can be considered stable. That only comes from an educated and alert furnace department that appreciates the impact they can have on the stability of the forming process.

Glass conditioning is perhaps one of those areas that is understood by the least amount of people in the glass industry, yet it has a major impact on product quality and productivity. This being the case, it will be welcome to see more work being done in this area. Not least of which should be the education of the furnace department, as well as the production department, in the role the forehearth plays and the limitations in control capability it may have.