OCR Text |
Show CONCLUSIONS The above case studies demonstrate that high order, nonlinear modeling tools such as Process Insights can build accurate emission models for a variety of unit operations that require emissions monitoring. These models, when coupled with Sensor Validation System to create a PEMS, provide highly accurate and reliable emissions monitoring. A PEMS has several advantages including: 1. The initial cost of installation and certification of PEMS is 1/3 to 1/2 the cost of a CEMS. The installation costs through certification costs of a PEMS is $70,000-$125,000 in comparison to $150,000-$300,000 or more for a hardware CEM. 2. PEMS has proven to be as accurate as the best available established monitoring. 3. The maintenance costs of PEMS are small in comparison to those of a CEMS. The maintenance costs for a PEMS is $10,000-$20,000 in comparison to $50,000 to $80,000 for a hardware CEM. 4. The Sensor Validation System provides PEMS with minute-by-minute quality assurance checks and superior reliability in comparison to a CEM. 5. PEMS satisfies the regulations of 117.580( c) recently passed in Texas for PEMS. 6. Perhaps most importantly, nonlinear regression modeling tools can provide very valuable information on the plant operations. CEMS can only provide information on whether the emissions are within the regulations or not. In contrast, PEMS can provide information on how to operate the process cmd - optimize the process to achieve maximum output and minimum emissions. PEMS provide for increased profits and a cleaner environment. References [1] F. Curtis, US EPA, Emission Measurement Center, May 4, 1995, letter [2] D. Rumelhart and J. McClelland, 1986. Parallel Distributed Processing, Cambridge, MA: The MIT Press. [3] Advances in Neural Information Processing Systems 1-5, 1989-1993. San Mateo, CA: Morgan Kaufmann Publishers. [4] J. Hertz, A. Krogh, and R. Palmer, 1991. Introduction to the Theory of Neural Computation. Addison-Wesley. [5] V. Rao Vemuri, 1992. Artificial Neural Networks: Concepts and Control Applications. Los Alamitos, CA: IEEE Computer Society Press. 13 |