Brewery and power station pilot sites evaluate energy efficiency technology

ICI's Thornton Power Station in Lancashire and the Carlsberg Tetley brewery in Burton on Trent are currently testing the potential of advanced data analysis to reduce levels of energy consumption. The pilot projects are part of the UK Department of Environment's Energy Efficiency "Best Practice" programme and use XpertRule data mining software from Lancashire-based XpertRule Software Limited.

Patterns of energy use are complex in process industries, where it is difficult to understand the events that cause energy usage to rise and fall, particularly when production rates are highly variable, when product mix varies, or when there are several interacting processes on a single site. Data mining techniques are used to analyse the factors that affect efficiency and models are automatically generated to improve decisions taken, for example, about the best combinations and loading of plant or about the frequency of cleaning heat exchangers.

The first phase of the project, which has just been completed, aims to demonstrate the capability of Rule Induction, Genetic Algorithms and Neural Networks to assess and optimise energy usage in industrial processes. The second phase involves development of an easy-to-use software package which is then hoped to encourage wider use of the technology by process engineering and management teams.

Carlsberg Tetley Burton Brewery employs seven refrigeration compressor units as part of the production process. with an annual energy consumption of over 10 million KW Hours, refrigeration accounts for 30% of the plant's energy usage, which has an annual cost of around UK £1.8 million.

Tony Sykes, Chief Engineer at the brewery, comments on the project: "We have sophisticated, site-wide supervisory systems that provide detailed, real-time data. The problem is that we have masses of information which makes it difficult to analyse performance of the refrigeration plant and isolate the factors that will help to achieve the best levels of performance and energy efficiency. It is a quite different approach to that required for equipment maintenance." He adds: "The project involves gathering information about performance and energy usage and doing something quickly with it to achieve improvements and reduce costs. We are keen to be at the sharp end of best practice in this area."

A refrigeration system optimiser takes account of the efficiencies of the various compressors, condensers and evaporators while anticipating the future demand for cooling and the effect on costs of varying electricity prices. A rule-based monitoring system has also been developed for the refrigeration system which uses examples of previous faults to identify new faults quickly and reliably without the need for an expert on site.

Thornton Power Station was commissioned in 1958 to supply steam for process heating and electricity for ICI's entire chemical manufacturing Hillhouse site. The station has three 98 MW thermal input high pressure water tube boilers, as well as three steam turbines, and the annual cost of natural gas and heavy fuel oil runs well into seven figures. Although new instrumentation and control equipment has replaced existing systems, efficiencies from the dated plant fall far short of modern gas fired combined heat and power installations.

Stuart Barlow, Operations Manager at the power station, comments: "Past experience with similar technology persuaded me that it was worth exploring the use of advanced data analysis, initially targeted at the station's boilers. A small percentage saving would have a significant effect on the energy costs and this project has the potential for realising considerable benefits by using these new techniques."

The cost of software tools used in the project is expected to be around £10,000, with implementation costs varying, depending on the type of data interfacing routines required to connect into existing computing and control systems. It is anticipated that the system will be suitable for all medium and large sites in process industries for whom the system is likely to achieve a payback in one or two years, based on energy savings alone.

About the "Best Practice" programme

• The limited understanding of the potential benefits of energy efficient technologies and techniques
• The lack of objective information on both existing and novel energy efficient technologies and techniques
• Institutional barriers, particularly within the building sector
• The weak and fragmented nature of parts of the energy efficiency industry