Programme

October 14th

The use of AI in Industrial Facilities  such as Power Generation, Oil & Gas, Processing and Chemicals is not new; in fact machine learning techniques for monitoring the condition of critical assets (Turbines, Compressors, Boilers) has been around for almost 20 years. Advances in AI techniques, however, such as language models and reinforcement learning bring new possibilities to the way we design, operate and optimise these high integrity facilities to accomplish higher efficiency, lower emissions, greater productivity and autonomy.

AVEVA is a global leading Industrial Software Company, founded in Cambridge and develops software for the design and operation of some of the most complicated industrial facilities. Dr David Smith is a chartered Mechanical Engineer and spent the first half of his career in Industry working for Power Generation companies. At AVEVA he leads the application of new emerging AI technologies to solve Industrial Problems. This talk will focus on the application of AI technologies to accomplish advances in the way we design, operate and maintain large industrial facilities. How we can use telemetry data from critical equipment together with AI, to catch issues which could lead to £ millions of lost production, How can we use AI techniques to perform routine tasks in design like pipe routing, and even how we can build an AI agent capable of controlling a facility through difficult operations that would take a human operator years to learn.

• Brief description of signalling systems down the ages
• Purpose and Principles
• Block sections and signal aspects
• Mechanical signals
• Colour light signals
• Detection of trains
• Route signalling
• Points
• Interlockings
• Train describers
• Single lines
• Level crossings
• Train protection systems
• Modern era – ERTMS and ETCS
• Summary

The establishment of a network of telegraphs allowed the collection of weather observations in real time. This led to the issuing of the first public weather forecast in 1861. Since then technological developments and enhanced understanding of atmospheric processes have underpinned the increase in accuracy of weather forecasts. In particular, from the mid-1960s the use of increasingly sophisticated numerical models coupled with an enhanced availability of observations, especially using satellites, have led to modern five-day forecasts being as accurate as one-day forecasts 50 years ago. These developments will be discussed along with how numerical models are used to investigate climate change.

Bob Riddaway After gaining a BSc in Physics and PhD in Meteorology from Edinburgh University, he joined the Met Office to do research. He soon found that training and operational meteorology were more to his liking, so his career included being Head of London Weather Centre, Head of Training, Joint Head of Forecasting and Head of Development Resourcing & Technology. Whilst at the Met Office he became involved in the education and training activities of World Meteorological Organization (WMO). After retiring from the Met Office his involvement with WMO continued and he also worked as a consultant at the European Centre for Medium-Range Weather Forecasts (ECMWF) for ten years. For nearly fifty years Bob Riddaway has been involved in the activities of the Royal Meteorological Society (RMetS). These included running various educational activities, helping to establish professional meteorological qualifications and being the founding editor of the journal Meteorological Applications. For eight years he was the RMetS General Secretary and is currently a member of the Accreditation Board. He has also been the President of the European Meteorological Society (a society of meteorological societies).

Roger will open his talk with a brief review of his career covering the design and commissioning of large power systems – in particular a 360MW power station in Malaysia. He will take time to explain certain technical terms (such as Synchronous/Asynchronous operation, System Inertia, and other terms mentioned below). Moving on to the Government’s target for NZ (Net Zero carbon release to the atmosphere) he will highlight the UK’s lack of money, materials (and associated lead times), and skilled resources to meet said Target. This will be illustrated with particular reference to: 1. Providing spinning reserve power, to meet Maximum Demand (MD) – (including droop control) – by the target date of 2035. By then around 40 GW of aging nuclear and gas powered generators will have been retired. 2. The portion of power capacity provided by ‘Renewables’ that must be backed up by non-Renewables to cater for sustained periods of calm, cloudy, cold UK (winter) weather. 3. Interconnectors (to other countries); capacity and availability. 4. More subtle engineering considerations such as: Rotational Inertia (aka Generator Stability), MVAR capacity, and fault clearance capability. 5. The role of Grid scale batteries and potential Hydrogen usage and storage. 6. Up to date Grid control to deal with Grid intermittency. 7. Sufficient distributed Black Start capability. (The RCEA visits to Shoreham Power Station touched on this.) Such factors rudely came to the fore during the recent Iberian Blackout, and may well do so again if (and when?) a similar situation arises in the UK.

We continue to be told that wind and solar will be free, when the cost of electricity is already 30% higher than the wholesale price, (due to green levies) and wind capacity factors have been reduced by almost 30% relative to those quoted in the original NZ case for their use. Also the cost of back up capacity (see bullet 2 above) must be included when costing Renewables. The Royal Society estimates that 100TWh of energy will be needed pa for that Hydrogen is also cited as an alternative green energy storage medium but the technology is relatively new, expensive, and carries other challenges. The ongoing availability of nuclear fuel has also been cited as an issue: Roger will touch on the under used possibilities of re-processing spent fuel. Added to all this is the cost of the extra distribution lines and equipment. NG (National Grid) estimated £3trn, just for the U.K. grid upgrading. Then we need to add in costs of local network upgrading as well as potential subsidies for heat pumps, home insulation etc. McKinsey’s (a well respected independent consultancy) estimated the TOTAL cost of UK Net Zero at £5 trn, which they equated to 7% of GDP pa through to 2050. However, DESNZ continues to quote <£1trn and, worryingly, say that they don’t have a viability analysis as such, for Net Zero.