Preface
Virtually all the work in the Science and Engineering Section, and in the Economics and Industry Section has been done with the support and collaboration of other people. Their effort is acknowledged below and in more detail as joint authors of the published and private work listed in the sections themselves.
This has come about because after about two and a half years after joining ICI’s Bozedown Laboratory from Cambridge as a Technical Officer in 1964, when Graham Neilson became my Laboratory Assistant, I became responsible for a section of 6-9 other Technical Officers (graduates and PhD’s) each of whom had a project with supporting technical staff. After a further three years, I was appointed head of the Process Technology Group of about 25 Technical Officers and 30 supporting staff, together with direct access to a substantial workshop and the Company’s KDF9 digital computer at Wilton on Teesside (North Yorkshire). I reformed the Group into five project teams with about five Technical Officers each, covering the following fields: Gas-solid catalytic reactors (Dr M L Brisk, later Professor of Process Control & Dean of Engineering, RMIT); Free Radical computation kinetics (Dr P Dyer); Gas-liquid reactors (Dr C Ramshaw, later Professor of Chemical Engineering, University of Newcastle); Two-phase & polymer reaction flows (M J Shires, later Senior Engineer with Foster–Wheeler); Gas-phase reactors (self). The scope and scale of the five teams made the Process Technology Group (PTG) the largest chemical technology research group in Imperial Chemical Industries (ICI), which in the 1960s was the largest chemical company by turnover in the world, and still in the first five two decades later.
Each of these teams was aimed at reactor design and process improvement on the full-scale through the basic philosophy of PTG: separate out the scale-dependent from the scale–invariant factors.
Scale-dependent factors included fluid flow, mixing and heat transfer; scale-invariant factors included chemical kinetic constants, thermal and mass diffusion constants, droplet formation kinetics. The scale-dependent factors themselves included scale invariants such as viscosity and thermal conductivities. Separating these factors, measuring them on the lab scale, then combining them into mathematical models, which could then reliably be applied to the factory scale, was the major achievement of PTG, to which all members contributed.
The theory of doing this with a greatly extended range of process is the basis of the Science of Process Manufacture (qv) and the lecture course given by the writer for 10 years in UMIST to the end of 2005. These can be found in “University Lecture Courses” as part of the UMIST Polymer & Mechanical Engineering (1979-99) section.
C A J Young
Christopher (“A J”) Young, an Oxford Physicist by training, was recruited by Sir Ewart Smith FRS, then ICI Engineering Director, from the Sudan Meteorological Service in 1946 and was asked to set up a small Instruments Section at the Frythe Laboratories in Hertfordshire. In 1956, the Control and Instruments Section as it had become was moved to Bozedown House, Oxfordshire, under “A J” as Group Manager, reporting directly to the ICI Main Board Director. There were 3 Section Managers: Ivor Gray, Douglas Whiting and Ray (R L) Day.
GB Patent 1,422,781 28th January 1976; US Patent 3,969,449 13th July 1976.
M J Shires and S F Bush
Abstract
A heat sensitive liquid is defined as a liquid which degrades, usually to carbon, at temperatures below its normal vaporisation temperature. This patent* describes an invention whereby a heat sensitive liquid is vaporised with exceptionally low thermal degradation.
Many such liquids, for example light naphthas, used to make olefines, are vaporised in tubes in which the liquid passes from pure liquid to pure vapour through a two-phase vapour core-liquid film flow regime (qv). Because the organic liquid droplets are always present in the vapour, they impinge on the hot tube walls downstream of the last of the liquid film, degrade to carbon and then by diffusion into the steel tubes making them brittle and appreciably shortening their lives, with all that implies in terms of furnace shutdown and repairs.
While hydrocarbon liquids do not degrade appreciably to carbon below cracking temperatures (800+ oC), hydroxyocarbon liquids degrade at much lower temperatures, typically 250 oC. An important example of this is the dicarboxylic adipic acid (CH2)4(CO OH)2 which is needed to be mixed with ammonia NH3 in 1:2 molar proportions over a solid catalyst to make adiponitrile (CH2)6(NH2)2 the last step before polymerisation with adipic acid to make nylon poly(hexamethylene amide) – in effect a synthetic protein with a single repeating acid group. All these processes we carried out on the largest industrial of scales, typically 2-3 tonnes per hour.
Carbon and tarry residues formed at the adipic vaporisation stage have deleterious effects on the solid catalyst at the adiponitrile stage. Shortening its life, requiring around 700 tubes to be recharged at often monthly intervals, causing excessive lost production time and cost of catalyst replacement.
The invention provides a general principle of design by which droplets of liquid do not impinge to measurable degree on heating surface above the degradation temperature in the adipic acid case 250 oC which is some 60 oC below the one atmosphere boiling temperature. This is achieved by combining two technologies:
- liquid (adipic acid) atomisation and
- internal jet flows (qv) produced by ammonia gas needed for the next, synthesis of adiponitrile, stage.
The patent describes how the ammonia jet in a relatively compact vaporiser (about 1.3 metres in diameter for 2 tonnes per hour of adipic acid) creates a circulation rate of some 20-30 times the ammonia flow, such that unvaporised adipic acid droplets from their entry (atomisation) point at 180 oC are never in the vicinity of the heating surfaces (typically 450-480 oC).
This innovation demonstrates again the versatility of the jet principle not only in process engineering, but also in the design of hydrocarbon combustion proceses, especially internal combustion chambers which also face ever-tightening demands to reduce carbon and tarry particulate emissions (see papers on diesel engine combustion, by Bush, Chen and Whitehouse).
Footnote
* These two patents are basically the same although GB 1,422,781 describes two alternatives.
To see the US patent in full, go to US Patent Office. On the first page click on USPTO Patent Full-Text and Image Database (PatFT), then under the heading “Searching Full Text Patents (since 1976)”, click on Patent Number Search and enter the patent number (with or without commas) into the “Query” box, then click on “search”. To search for another US patent, click on Pat Num in the red display at the top of the page.
Group II Research Note, ICI Corporate Lab.
S F Bush with P Dyer and M J Shires.
Summary
Experimental results from a new device for vaporizing heat sensitive liquids are reported. Experiments have been carried out with adipic acid on a 20 lbs/hr rig at CL-B. The direct contact of liquid adipic acid and heating surfaces is minimised by injecting acid droplets into an internally circulating vapour flow, which itself receives heat from the walls of the vaporizer. The internal circulation is maintained by entrainment into an incoming ammonia jet. Some experimental optimisation of the internal geometry has been achieved.
This principle of jet-spray vaporization is designed to be directly applicable in the existing spinner shells in Petrochemicals Division with only a minimum of modifications. It should also apply to other vaporizations where a carrier gas is available or can be introduced.
See also the patents arising from this work Vaporisation Process.
Group II Research Note, ICI Central Instrument Research Lab.
S F Bush with M J Shires and M A L Sita-Lumsden.
Summary
Results obtained from an examination of various aspects of the PTA process are given. These include esterification rate measurements, CP5 residence time measurements, an elucidation of the thin film results, a design of pre-polymeriser coil, a relationship between throughput, revolution speed and plate spacing in the CP5 design and an outline of a first model of the batch esterifier.
Group II Research Note, ICI Central Instrument Research Lab.
S F Bush with M J Shires.
Summary
The data enclosed is a summary of the two main alternative designs put forward by CIRL for a prototype unit to vaporise 7500 lbs/hr of adipic acid in the presence of an 8 : 1 molar ratio of ammonia to adipic acid.
