July 2017 – The Hobart Chinaman

What’s a deckle?
Here’s a definition. “Deckle a device in a papermaking machine for limiting the size of the sheet.”
Papermakers wish to maximise the use of the deckle when making gypsum board liner to maximise their productivity. The gypsum board makers want to limit the amount of linerboard they use.
Gypsum board is a sandwich of processed plaster between two layers of linerboard. A creamy face linerboard and a greyer, cheaper back linerboard. For the three common widths of gypsum board produced, there are six linerboard widths.
When producing gypsum board the wider face linerboard edges are wrapped over into an envelope to which the back linerboard is glued.
All linerboard is made on the same paper making machine. Its width, deckle is fixed. At the end of the paper making machine the deckle is trimmed to the widths to suit the gypsum board maker. The trims are wasted and recycled into the paper making process. Increasing amounts of recycle degrades the quality of the paper being made, so only so much can be recycled. The rest is waste.
It was a conundrum.
The gypsum board maker forever attempting to reduce the widths of face and back linerboard reducing the tonnages they purchased. The paper maker wasting more and more of the deckle width.
What if the paper maker could use the full deckle width?
What if the cut pieces were an exact divisor of the deckle widths?
I calculated that for face liner this would be three pieces and back liner four pieces
Further I calculated that with just two liner widths, a single face width and a single back width, all three common widths of gypsum board, 900mm, 1200mm and 1350mm could be produced!
There would be reduction in the sizes stocked at both producer and manufacturer. With the reduction in waste at the paper maker would only have to make liner for the gypsum board maker twelve times a year and not thirteen, a significant saving in trees and setup costs.
I arranged for the paper maker to make a face liner and back liner at the new special widths. We made a short run of gypsum board. The gypsum board produced had superior performance characteristics especially in its edge formation. The glued overlap of the envelope was further into the back of the sheet increasing strength.
The paper makers were at first intrigued and then enthusiastic. In fact, spread across the whole gypsum board making industry there were multi-million dollar savings to be had.
The paper maker proposed the savings to be made would be shared between the papermakers and gypsum board makers. The paper maker rightly pointed out that a change in paper widths was not protectable by a patent so the innovation might be requested by other gypsum board makers. They were agreeable though to recognizing the innovative nature of the idea.
However, the organisation I worked for was unable to see why the benefits should be shared for the whole industry and the country. They refused to progress the matter.
Of the two pallets of gypsum board made just a small sample remains somewhere in my treasured possessions.

The superphosphate plant ran unreliably, only 38% of available time. Its quality of output was poor and highly variable. It stopped and farted without apparent reason due to unexplained blockages. Raw material [calcium phosphate] came from around the world. Vietnam, sub- Saharan Africa, Christmas Island and Nauru. So its quality varied widely. I decided I needed to understand why the plant stopped so frequently and unexpectedly.
Two years of data ended up draping the walls of my office. The data included size of final product granules, temperature variations in the drying process, and amount of sulphuric acid mashed into raw ground material.
The raw calcium phosphate rock was milled in a combination train of six 20 tonne per hour Lopulco Mills, and a large 60 tonne per hour Ball Mill. To maintain the fineness and production rate targets, variable combinations of Lopulco and Ball Mills were used. Depending on which Mills were under repair, the output fineness of the combination was maintained at a fineness of 95% < 200 mesh sieve.
I spent nearly a month combing through the data correlating material combinations, output charts and delays. Which part of the data matched closest to the plant stoppages?
I tracked back into the process, stopping where we were unable to exert control.
The data showed that changes in the combination of mills correlated with subsequent delays, even though the fineness of the output was being maintained.
I realised that the reactivity of sulphuric acid and the phosphate rock was inconsistent. From my time in gypsum board manufacture I recalled that chemical reactivity is a function of particle size surface area and not fineness per se.
Taking samples of phosphate rock ground to the same fineness in the Lopulco Mill and the Ball Mill I had both samples tested at the chemical laboratory, where thirty years earlier I had begun my chemical engineering career.
Shazzam! The Ball Mill sample produced four hundred percent more surface area than an equivalent Lopulco Mill sample. The Ball Mill sample thus had much superior reactivity.
So the changes in mix of Mills dramatically affected the suitability of the ground material fed into the process.
Not long after suggesting operating specifications based on surface area, the plant achieved the best ever recorded daily, weekly and monthly production rates ever achieved.
The CEO of the organisation came along to present me with the Award for best operational performance worldwide.