Predictive Emission Monitoring Systems (PEMS) applications in energy and process industries

Affordable, accurate, compliant machine-learning-based solution for emission monitoring in oil & gas, chemicals, steel, pulp & paper, mining, cement and power utilities using gas turbines, boilers, furnaces, internal combustion engines and other complex process units. It requires the constant quality combustion fuel, as well as proper instrumentation and automation.

Environmental regulations are increasing, demanding at least 97.5% availability of emission monitoring solutions. This increases compliance costs, requiring 24/7 monitoring of pollutant gases in energy-intensive industries. Non-compliance has never been more expensive, in terms of both fines and threat of shutdown. Due to complexity of industrial operations, it is often difficult to to find  faults at the right time to avoid major shutdown.

Significantly reduce CAPEX & OPEX for regulatory compliant tracking of emissions released into the air

Contact Us

Predictive Emission Monitoring System (PEMS) represents a cost-effective approach for continuous monitoring of emissions as alternative to hardware-based Continuous Emission Monitoring Systems (CEMS). Since PEMS are software-based, they do not need gas analyzers and associated hardware like sample conditioning or shelters. Interfaced with plant control systems, PEMS utilize process inputs to offer continuous, real-time monitoring of pollutants, e.g. NOx, SO2, CO, HC or diluents like O2. PEMS are generally suitable for emission sources such as oil, natural gas, pulverized coal, coke in lieu of CEMS, providing equal accuracy and data quality:

  • primary emission monitoring platform when CEMS hardware is not feasible or cost effective
  • secondary emission monitoring solution backing-up CEMS hardware to increase infrastructure availability during failure and maintenance period
  • advisory and simulation tool to help plant operators run what-if analysis, simulating emission behavior at varying process conditions

PEMS exploit advanced mathematical models, which use process parameters (e.g. pressure, temperatures, flow, etc.) as input variables in order to estimate the most common air pollutant emissions (e.g. NOx, SO2, CO, CO2, etc.). They are:


~50% lower initial investment, operational expenditures and no unexpected costs due to unpredictable failures in the emission monitoring infrastructure


99.5%+ service uptime: software extends availability of emission data while maintaining equal accuracy and data quality as CEMS hardware-based solution


ABB's proprietary software for environmental applications under US-EPA and EU frameworks is a state-of-the-art and future-proof machine learning based solution.

PEMS for industrial steam & power plants in pulp & paper, metals, cement, mining, etc.


  • Boilers to feed steam turbines and or generate steam for production process
  • Gas turbines for captive power generation
  • Constant quality fuel source

PEMS for specific industrial processes

Pulp & paper mills
  • Recovery boiler of kraft process (combustion based on wood or natural gas)
  • Lime kiln in kraft process (if using stable energy as fuel oil or natural gas)

Metals plants
  • Reheat furnace (combustion based on pulverized coal)
  • Coke oven plant (combustion based on coke)

  • Crude oil pre-heating furnace
  • Propane compressor driving gas turbine

PEMS are compliant and accepted by international environmental regulations and standards.


  • U.S. EPA defines within 40 CFR 60 and 75 the procedures to use PEMS as alternative technology to HW CEMS
  • Performance Specifications 16 prescribes the specific requirements to check and certify model performances


  • Europe has released a comparable set of specifications to be applied for PEMS implementation
  • Technical Specification (TS 17198) defines PEMS requirements as back up and as alternative of conventional CEMS
  • The new standard is largely based on the NTA 7379 available in the Netherlands since a few years

Middle East

Several other countries are introducing PEMS within their environmental regulations: Middle East, Malaysia, etc.
ABB installations
Inferential Modeling Platform

Related insights

Loading documents
Select region / language