An upgrade of the Capitol Cement precalciner kiln line in 2000 resulted in increasing production by 33 percent. Clinker production outstripped the finish milling capacity to grind it to cement. In order to fully recognize the benefits of increased clinker production, the plant needed to eliminate the bottleneck created by a limited milling system. When Capitol determined that the required increased capacity would not justify the capital expense for a new mill or extensive modifications, the company searched for a solution that would use the existing system and controls. Pavilion Technologies proposed a solution that appeared to satisfy these conditions. The Process Perfecter ® has proven to be effective in increasing mill production and improving product quality by changing process conditions.

Neural network (NN) advanced process control technology was deployed quickly and successfully on a very large HDPE loop reactor. Process modeling was achieved in just a matter of weeks. Once NN controls were on-line, operator acceptance occurred almost immediately. Keys to success include a data-rich environment from process historians, the right NN technology, full coverage of all disciplines by in-house and vendor personnel possessing both expertise and enthusiasm, and a thorough safety analysis to allow deployment while the reactor was in full operation.

The presentation draws from 10 years of learning and experiences to address issues relating to the introduction of successful Pavilion advanced process control (APC) solutions into a large, decentralized New Zealand Company. Emphasis will be placed on the value proposition and the system, infrastructure and data requirements that enable realization of the benefits. In addition, the paper explores methodologies and techniques employed for enhancing user acceptance of the technology.

Spray drying is a thermodynamic process in which extract is transformed into a dried product by spraying it into a hot drying medium.

In order to improve the control of the major process variables, Nestlé began investigating an implementation of the Process Perfecter® in the coffee spray drying operations in 2001. The main target was to reduce fluctuations in the product moisture, density and color with a maximized production rate.

After trials in selected pilot and production plants, Nestlé's operations have noticed improvements in the process stability and reduction of operator interventions. Thus, the main benefits of implementing such technology will be to provide support to production people and obtain product consistency rather improving the performance.

The European IPPC (Integrated Pollution Prevention Control) directive is a framework for pollution control from all the major industries in the European Union. Axess Energy and Environment Ltd. has succeeded in getting PEMS implemented into the Best Available Technology (BAT) Reference document for NOx monitoring at Large Combustion Plant. This document will be used by the pollution authority in each Nation in approving BAT for each specific plant. PEMS as a monitoring technique will significantly reduce the monitoring costs for the industry, especially for offshore installations. The OPPC bureau has been very co-operative with the industry to seek cost effective solutions.

Teepak has developed some of the licensed steady-state tools online almost as far as any known Pavilion customer. The Pavilion-based optimization models have been running in closed-loop control for over a year. In the deployment multiple challenges were overcome including not only use of FANN models online and dynamic filtering for long process dynamics, but also use of a random number generator to interface with standard DCS watchdog timer functions. Beyond this, modeling solutions to managing correlated input data, inverted gains and advanced data filtering were used to develop good process models.

The NOVA Chemicals Corporation Joffre gas phase polyethylene plant has been a top producer of LLDPE for almost 20 years. This performance achievement has been supported by advanced implementation of resin property control, as well as multivariable production rate control and optimization. In April 2004 a project was approved to replace this control scheme with a Pavilion non-linear APC solution. This paper will review the business justification, project execution and results validation.

The SABIC EuroPetrochemicals plant in Gelsenkirchen, Germany produces 500,000 mt of polypropylene per year. The variety of different products produced in one plant over a wide range of conditions is a big challenge with transitions between different grades as low 2nd class production as well as staying consistent on stable conditions when producing at-grade. The main task of Pavilion Process Perfecter was to minimize 2nd class material during transitions and increase product consistency during the entire production circle. Deploying Process Perfecter enabled a new transition philosophy which increased first class production by 4.5% and in addition eliminated propylene venting of 300 mt per year. Furthermore product consistency increased by 30%.

Several advanced process control (APC) projects have been successfully implemented at the Sterling's plant. However, how does one measure success on a project like APC? This leads all the way back to justifying the APC for capital approval. In getting the APC for capital approval, one must assess the process-related benefits of APC and then be able to translate the process benefits into economic benefits that the approval managers can understand. These economic benefits would then become the objective criteria for measuring the success of the project. This paper will touch on several practical methodologies used in estimating APC benefits from an economic approach with examples.

Neural network models, once built, are fixed in time. However, the process to which they continually operate against in real-time changes continuously. This presentation discusses a process that can be used to maintain model life for long periods. Some examples used are for prediction applications that have been running continuously for six years and longer.

Process Perfecter Model Predictive Control Software prioritizes and maximizes feed to the Crude Unit of Nova Chemicals Corunna Site. The feed comes from two condensate and two crude tanks that receive batches from multiple sources around the world. Virtual on-line Analyzers give on-line estimates for product properties since lab analysis data are available twice daily. The estimates are updated as lab results become available. The control application moves product draws and tower temperature to maintain products to specification.

As olefins business managers around the globe adjust operating plans for the next five years, questions are being asked about a variety of strategic issues facing the industry today and those that are likely to emerge near term. What impact will high crude oil prices have on global economic growth and therefore on light olefins demand growth? How will new olefins unit startups in Asia and the Middle East impact the balance of supply and demand? What will set competitive olefins economics in the future? What will be the impact of high natural gas prices in North America? Under what scenario will light olefins markets begin to turn down? This presentation will provide our views regarding these questions and present the audience with a basis for reaching their own conclusions on these issues.