Situation: Industrial facilities, such as those used for power generation, waste-to-energy, chemical processing and refining, are often required to monitor air emissions such as nitrogen oxides, carbon monoxide, carbon dioxide and sulfur dioxide. Predictive emissions monitoring systems (PEMS) offer one solution for ensuring regulatory compliance. PEMS use computer models to relate the inputs of the combustion process to the emissions produced in order to continuously predict stack emissions. nDimensional’s client provides a range of compliance management systems, including PEMS, for hundreds of clients across multiple industrial sectors. However, their PEMS offering used costly third-party computer models over which they had little control. They engaged nDimensional to assist them in developing their own in-house solution.
Solution: The nD platform was used to develop and train predictive emissions models using a combination of domain expert algorithms and advanced machine learning. The models were tested and validated to ensure high precision and packaged with an easy-to-use interface that integrates seamlessly with the client’s web-based data acquisition and handling system. This new PEMS system combines the latest modeling technology with existing reliability, security, and reporting capabilities to ensure full regulatory compliance. It enables our client to deliver a complete PEMS solution at lower costs and higher margins.