Provide PDF Format
PRCI PR-258-9619
- Pressure Based Parametric Emission Monitoring Systems (PEMS)
- Report / Survey by Pipeline Research Council International, 04/02/2002
- Publisher: PRCI
$348.00$695.00
L51916e
Enginuity International, Inc.
Need: The natural gas industry operates over 8000 stationary large bore (bore > 14 in) natural gas engines for natural gas compression on pipelines and power generation. As emissions regulations become increasingly more stringent, the need for low cost methods for compliance demonstration arises. A PEMS model is one such approach. Research in this area has increased significantly during the last decade. PEMS models for this application utilize parameters commonly measured on industrial engines in the field to predict engine-out emissions. Monitoring emissions in this manner represents a significant cost savings over the periodic use of chemiluminescence NOX analyzers, which are not standard equipment in natural gas compressor stations. PEMS model accuracy is dependent on the quality of the input data, both the training NOX measurements and the selection of input parameters. Hence, it is important to have both reliable data measurement methods and an understanding of engine operating parameters relation to NOX.
Result: This work is part of the body of work referred to as the Integrated Test Plan (ITP), performed at the Engines and Energy Conversion Laboratory (EECL). This report details an investigation into Parametric Emissions Monitoring System (PEMS) models. It is the final document to be delivered under the ITP program. Much of the work performed under the ITP program focused on Hazardous Air Pollutants (HAPs) research. However, the emphasis of the PEMS work is on the prediction of oxides of nitrogen (NOX) emissions from large bore natural gas engines.
Benefit: In this work two different PEMS models are developed, a semi-empirical model and a neural networkmodel. The semi-empirical model is based on general relationships between NOX emissions and engine parameters, but contains empirical constants that are determined based on the best fit to engine experimental data. The neural network model utilizes a similar set of input parameters, but relies on the neural network code to determine the relationships between input parameters and measured NOX emissions. The neural network model also contains empirical constants. The mathematics involved in both models is described. A single term semi-empirical model, which has been utilized in the literature as a PEMS model, is applied for comparative purposes.
Enginuity International, Inc.
Need: The natural gas industry operates over 8000 stationary large bore (bore > 14 in) natural gas engines for natural gas compression on pipelines and power generation. As emissions regulations become increasingly more stringent, the need for low cost methods for compliance demonstration arises. A PEMS model is one such approach. Research in this area has increased significantly during the last decade. PEMS models for this application utilize parameters commonly measured on industrial engines in the field to predict engine-out emissions. Monitoring emissions in this manner represents a significant cost savings over the periodic use of chemiluminescence NOX analyzers, which are not standard equipment in natural gas compressor stations. PEMS model accuracy is dependent on the quality of the input data, both the training NOX measurements and the selection of input parameters. Hence, it is important to have both reliable data measurement methods and an understanding of engine operating parameters relation to NOX.
Result: This work is part of the body of work referred to as the Integrated Test Plan (ITP), performed at the Engines and Energy Conversion Laboratory (EECL). This report details an investigation into Parametric Emissions Monitoring System (PEMS) models. It is the final document to be delivered under the ITP program. Much of the work performed under the ITP program focused on Hazardous Air Pollutants (HAPs) research. However, the emphasis of the PEMS work is on the prediction of oxides of nitrogen (NOX) emissions from large bore natural gas engines.
Benefit: In this work two different PEMS models are developed, a semi-empirical model and a neural networkmodel. The semi-empirical model is based on general relationships between NOX emissions and engine parameters, but contains empirical constants that are determined based on the best fit to engine experimental data. The neural network model utilizes a similar set of input parameters, but relies on the neural network code to determine the relationships between input parameters and measured NOX emissions. The neural network model also contains empirical constants. The mathematics involved in both models is described. A single term semi-empirical model, which has been utilized in the literature as a PEMS model, is applied for comparative purposes.
Related Products
PRCI PR-162-157
Effects of Random Seas on Pipeline Stability, Volume 1: Wave/Soil Simulations and Simplified Design ..
$6.00 $12.00
PRCI Report 178
Evaluation of an X80 Pipe Fabricated from a Plate Processed using On-line Accelerated Cooling..
$6.00 $12.00
PRCI PR-247-9511
Effects of Phosphoric Acid Treatment on Performance of FBE Coating..
$173.00 $345.00